1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
12 buffer overrun for (non-chunking) other transports.
14 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
15 TLS1.3, means that a server rejecting a client certificate is not visible
16 to the client until the first read of encrypted data (typically the
17 response to EHLO). Add detection for that case and treat it as a failed
18 TLS connection attempt, so that the normal retry-in-clear can work (if
21 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
22 and/or domain. Found and fixed by Jason Betts.
24 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
25 configuration). If a CNAME target was not a wellformed name pattern, a
28 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
29 the OS reports them interleaved with other addresses.
36 JH/01 Remove code calling the customisable local_scan function, unless a new
37 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
39 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
40 non-signal-safe functions being used.
42 JH/03 Bug 2269: When presented with a received message having a stupidly large
43 number of DKIM-Signature headers, disable DKIM verification to avoid
44 a resource-consumption attack. The limit is set at twenty.
46 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
47 report of oldest_pass in ${authres } in consequence, and separate out
48 some descriptions of reasons for verification fail.
50 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
51 files in the spool were present and unlocked. A queue-runner could spot
52 them, resulting in a duplicate delivery. Fix that by doing the unlock
53 after the unlink. Investigation by Tim Stewart. Take the opportunity to
54 add more error-checking on spoolfile handling while that code is being
57 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
58 No known attacks, no CVE, this is defensive hardening.
60 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
61 a queue-runner could start a delivery while other operations were ongoing.
62 Cutthrough delivery was a common victim, resulting in duplicate delivery.
63 Found and investigated by Tim Stewart. Fix by using the open message data
64 file handle rather than opening another, and not locally closing it (which
65 releases a lock) for that case, while creating the temporary .eml format
66 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
68 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
69 $sender_verify_failure/$recipient_verify_failure to "random".
71 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
74 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
75 Previously this would segfault.
77 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
80 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
81 like zero, since the resolver should be doing this for us, But we need one
82 as a CNAME but no MX presence gets the CNAME returned; we need to check
83 that doesn't point to an MX to declare it "no MX returned" rather than
84 "error, loop". A new main option is added so the older capability of
85 following some limited number of chain links is maintained.
87 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
89 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
90 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
91 now-deprecated earlier definitions used only specified the range up to TLS
92 1.2 (in the older-version library docs).
94 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
96 JH/15 Rework TLS client-side context management. Stop using a global, and
97 explicitly pass a context around. This enables future use of TLS for
98 connections to service-daemons (eg. malware scanning) while a client smtp
99 connection is using TLS; with cutthrough connections this is quite likely.
101 JH/16 Fix ARC verification to do AS checks in reverse order.
103 JH/17 Support a "tls" option on the ${readsocket } expansion item.
105 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
106 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
107 Previously the "utf8" would be re-prepended for every additional message.
109 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
110 Previously thery were accepted, resulting in issues when attempting to
111 forward messages to a non-supporting MTA.
113 PP/02 Let -n work with printing macros too, not just options.
115 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
116 one parent address was copied, and bogus data was used at delivery-logging
117 time. Either a crash (after delivery) or bogus log data could result.
118 Discovery and analysis by Tim Stewart.
120 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
121 Previously if the string ended mid-character, we did not insert the
122 promised '?' replacement.
124 PP/04 Documentation: current string operators work on bytes, not codepoints.
126 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
127 should pack well giving a smaller memory footprint so better caching and
128 therefore performance. Group the declarations where this can't be done so
129 that the byte-sized flag variables are not interspersed among pointer
130 variables, giving a better chance of good packing by the compiler.
132 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
133 non-null, to avoid issues with sites running BATV. Previously reports were
134 sent with an empty envelope sender so looked like bounces.
136 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
137 The ignore_error flag wasn't being returned from the filter subprocess so
138 was not set for later routers. Investigation and fix by Matthias Kurz.
140 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
141 and a msg:complete for the whole, when a message is manually removed using
142 -Mrm. Developement by Matthias Kurz, hacked on by JH.
144 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
145 a "Gnu special" function, asprintf() in the DB utility binary builds; I
146 hope that is portable enough.
148 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
149 requiring a known-CA anchor certificate; make it now rely entirely on the
150 TLSA as an anchor. Checking the name on the leaf cert against the name
151 on the A-record for the host is still done for TA (but not for EE mode).
153 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
154 included in delivery lines for non-proxied connections, when compiled with
155 SUPPORT_SOCKS and running with proxy logging enabled.
157 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
158 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
159 move the existing event to fire before the normal logging of message
160 failure so that custom logging is bracketed by normal logging.
162 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
163 msg:fail:internal event. Developement by Matthias Kurz.
165 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
166 far too small for todays use of crypto signatures stored there. Go all
167 the way to the max DNS message size of 64kB, even though this might be
168 overmuch for IOT constrained device use.
170 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
171 copy a string over itself. The library routine is documented as not
172 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
174 JH/32 For main options check_spool_space and check_inode_space, where the
175 platform supports 64b integers, support more than the previous 2^31 kB
176 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
177 the previous G, M, k.
179 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
180 $authenticated_fail_id variable on authentication failure. Previously
183 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
184 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
185 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
186 GNUTLS_SEC_PARAM_MEDIUM.
188 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
189 side. Previously we would continue as if no SNI had been received.
191 JH/36 Harden the handling of string-lists. When a list consisted of a sole
192 "<" character, which should be a list-separator specification, we walked
193 off past the nul-terimation.
195 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
196 causes) even when the retry time is not yet met. Previously they were
197 not, meaning that when (say) an account was over-quota and temp-rejecting,
198 and multiple senders' messages were queued, only one sender would get
199 notified on each configured delay_warning cycle.
201 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
203 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
204 cipher-suites, an error can be left on the stack even for a succeeding
205 accept; this results in impossible error messages when a later operation
208 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
209 return error codes indicating retry. Under TLS1.3 this becomes required.
211 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
212 it only wrote the new authenticators, resulting in a lack of tracking of
213 peer changes of ESMTP extensions until the next cache flush.
215 JH/41 Fix the loop reading a message header line to check for integer overflow,
216 and more-often against header_maxsize. Previously a crafted message could
217 induce a crash of the recive process; now the message is cleanly rejected.
219 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
220 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
226 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
227 When redis_servers is set to a list of > 1 element, and the Redis servers
228 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
229 case of MOVED into a DEFER case instead, thus moving the query onto the
230 next server in the list. For a cluster of N elements, all N servers must
231 be defined in redis_servers.
233 GF/02 Catch and remove uninitialized value warning in exiqsumm
234 Check for existence of @ARGV before looking at $ARGV[0]
236 JH/01 Replace the store_release() internal interface with store_newblock(),
237 which internalises the check required to safely use the old one, plus
238 the allocate and data copy operations duplicated in both (!) of the
239 extant use locations.
241 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
242 modifier. This matches the restriction on the commandline.
244 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
245 Previously only the last row was returned.
247 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
248 we assumed that tags in the header were well-formed, and parsed the
249 element content after inspecting only the first char of the tag.
250 Assumptions at that stage could crash the receive process on malformed
253 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
254 While running the DKIM ACL we operate on the Permanent memory pool so that
255 variables created with "set" persist to the DATA ACL. Also (at any time)
256 DNS lookups that fail create cache records using the Permanent pool. But
257 expansions release any allocations made on the current pool - so a dnsdb
258 lookup expansion done in the DKIM ACL releases the memory used for the
259 DNS negative-cache, and bad things result. Solution is to switch to the
260 Main pool for expansions.
261 While we're in that code, add checks on the DNS cache during store_reset,
262 active in the testsuite.
263 Problem spotted, and debugging aided, by Wolfgang Breyha.
265 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
266 When none of the hosts presented to a transport match an already-open
267 connection, close it and proceed with the list. Previously we would
268 queue the message. Spotted by Lena with Yahoo, probably involving
271 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
272 Previously a spurious "250 OK id=" response was appended to the proper
275 JH/08 The "support for" informational output now, which built with Content
276 Scanning support, has a line for the malware scanner interfaces compiled
277 in. Interface can be individually included or not at build time.
279 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
280 by the template makefile "src/EDITME". The "STREAM" support for an older
281 ClamAV interface method is removed.
283 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
284 rows affected is given instead).
286 JH/11 The runtime Berkeley DB library version is now additionally output by
287 "exim -d -bV". Previously only the compile-time version was shown.
289 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
290 SMTP connection. Previously, when one had more receipients than the
291 first, an abortive onward connection was made. Move to full support for
292 multiple onward connections in sequence, handling cutthrough connection
293 for all multi-message initiating connections.
295 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
296 routers. Previously, a multi-recipient message would fail to match the
297 onward-connection opened for the first recipient, and cause its closure.
299 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
300 a timeout on read on a GnuTLS initiating connection, resulting in the
301 initiating connection being dropped. This mattered most when the callout
302 was marked defer_ok. Fix to keep the two timeout-detection methods
305 JH/15 Relax results from ACL control request to enable cutthrough, in
306 unsupported situations, from error to silently (except under debug)
307 ignoring. This covers use with PRDR, frozen messages, queue-only and
310 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
312 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
313 metadata, resulting in a crash in free().
315 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
316 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
317 Broken also in d185889f4, with init system revamp.
319 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
320 Previously we abruptly closed the connection after reading a malware-
321 found indication; now we go on to read the "scan ok" response line,
324 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
325 ACL. Previously, a crash would result.
327 JH/19 Speed up macro lookups during configuration file read, by skipping non-
328 macro text after a replacement (previously it was only once per line) and
329 by skipping builtin macros when searching for an uppercase lead character.
331 JH/20 DANE support moved from Experimental to mainline. The Makefile control
332 for the build is renamed.
334 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
335 was allocated for every new TLS startup, meaning one per message. Fix
336 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
338 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
339 reported the original. Fix to report (as far as possible) the ACL
340 result replacing the original.
342 JH/23 Fix memory leak during multi-message connections using STARTTLS under
343 OpenSSL. Certificate information is loaded for every new TLS startup,
344 and the resources needed to be freed.
346 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
348 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
351 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
352 DATA response info to the (existing) per-recipient response info for
353 the "C=" log element. It can have useful tracking info from the
354 destination system. Patch from Simon Arlott.
356 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
357 length value. Previously this would segfault.
359 HS/02 Support Avast multiline protoocol, this allows passing flags to
360 newer versions of the scanner.
362 JH/28 Ensure that variables possibly set during message acceptance are marked
363 dead before release of memory in the daemon loop. This stops complaints
364 about them when the debug_store option is enabled. Discovered specifically
365 for sender_rate_period, but applies to a whole set of variables.
366 Do the same for the queue-runner and queue-list loops, for variables set
367 from spool message files. Do the same for the SMTP per-message loop, for
368 certain variables indirectly set in ACL operations.
370 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
371 as a multi-recipient message from a mailinglist manager). The coding had
372 an arbitrary cutoff number of characters while checking for more input;
373 enforced by writing a NUL into the buffer. This corrupted long / fast
374 input. The problem was exposed more widely when more pipelineing of SMTP
375 responses was introduced, and one Exim system was feeding another.
376 The symptom is log complaints of SMTP syntax error (NUL chars) on the
377 receiving system, and refused recipients seen by the sending system
378 (propating to people being dropped from mailing lists).
379 Discovered and pinpointed by David Carter.
381 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
382 replaced by the ${authresults } expansion.
384 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
386 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
387 allows proper process termination in container environments.
389 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
390 Previously the "final dot" had a newline after it; ensure it is CR,LF.
392 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
393 and "err_perm", deprecated since 4.83 when the RFC-defined words
394 "temperror" and "permerror" were introduced.
396 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
397 transport-filters or DKIM-signing. The restriction was lost in the
398 consolidation of verify-callout and delivery SMTP handling.
399 Extend the restriction to also cover ARC-signing.
401 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
402 in defer=pass mode supply a 450 to the initiator. Previously the message
405 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
406 tls_require_ciphers is used as before.
408 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
409 "pass_unscanned". Only tmpfails from the scanner are written to
410 the paniclog, as they may require admin intervention (permission
411 denied, license issues). Other scanner errors (like decompression
412 bombs) do not cause a paniclog entry.
414 JH/36 Fix reinitialisation of DKIM logging variable between messages.
415 Previously it was possible to log spurious information in receive log
418 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
419 triggered odd behaviour from Outlook Express clients.
421 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
424 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
425 since the IETF WG has not yet settled on that versus the original
426 "bare" representation.
428 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
429 Previously the millisecond value corrupted the output.
430 Fix also for syslog_pid=no and log_selector +pid, for which the pid
431 corrupted the output.
437 JH/01 Rework error string handling in TLS interface so that the caller in
438 more cases is responsible for logging. This permits library-sourced
439 string to be attached to addresses during delivery, and collapses
440 pairs of long lines into single ones.
442 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
443 during configuration. Wildcards are allowed and expanded.
445 JH/02 Rework error string handling in DKIM to pass more info back to callers.
446 This permits better logging.
448 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
449 do not close it down and have the child transport start it up again on
450 the passed-on TCP connection. Instead, proxy the child (and any
451 subsequent ones) for TLS via a unix-domain socket channel. Logging is
452 affected: the continued delivery log lines do not have any DNSSEC, TLS
453 Certificate or OCSP information. TLS cipher information is still logged.
455 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
456 identical IP addresses on different listening ports. Will also affect
459 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
460 add noisy ifdef guards to special-case this sillyness.
461 Patch from Bernd Kuhls.
463 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
464 than 255 are no longer allowed.
466 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
467 Disable the session-cache too, which might reduce our load. Since we
468 currrectly use a new context for every connection, both as server and
469 client, there is no benefit for these.
470 GnuTLS appears to not support tickets server-side by default (we don't
471 call gnutls_session_ticket_enable_server()) but client side is enabled
472 by default on recent versions (3.1.3 +) unless the PFS priority string
475 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
476 <https://reproducible-builds.org/specs/source-date-epoch/>.
478 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
479 the check for any unsuccessful recipients did not notice the limit, and
480 erroneously found still-pending ones.
482 JH/08 Pipeline CHUNKING command and data together, on kernels that support
483 MSG_MORE. Only in-clear (not on TLS connections).
485 JH/09 Avoid using a temporary file during transport using dkim. Unless a
486 transport-filter is involved we can buffer the headers in memory for
487 creating the signature, and read the spool data file once for the
488 signature and again for transmission.
490 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
491 4.77 as the kernel support then wasn't solid, having issues in 64bit
492 mode. Now, it's been long enough. Add support for FreeBSD also.
494 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
495 case where the routing stage had gathered several addresses to send to
496 a host before calling the transport for the first, we previously failed
497 to close down TLS in the old transport process before passing the TCP
498 connection to the new process. The new one sent a STARTTLS command
499 which naturally failed, giving a failed delivery and bloating the retry
500 database. Investigation and fix prototype from Wolfgang Breyha.
502 JH/12 Fix check on SMTP command input synchronisation. Previously there were
503 false-negatives in the check that the sender had not preempted a response
504 or prompt from Exim (running as a server), due to that code's lack of
505 awareness of the SMTP input buffering.
507 PP/04 Add commandline_checks_require_admin option.
508 Exim drops privileges sanely, various checks such as -be aren't a
509 security problem, as long as you trust local users with access to their
510 own account. When invoked by services which pass untrusted data to
511 Exim, this might be an issue. Set this option in main configuration
512 AND make fixes to the calling application, such as using `--` to stop
515 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
516 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
517 responses to those, into a single TLS record each way (this usually means
518 a single packet). As a side issue, smtp_enforce_sync now works on TLS
521 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
522 affects you only if you're dancing at the edge of the param size limits.
523 If you are, and this message makes sense to you, then: raise the
524 configured limit or use OpenSSL 1.1. Nothing we can do for older
527 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
528 cmdline element to get the documented default one. Previously it was
531 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
532 are unsuable for non-message-specific reasons - eg. connection timeout,
533 banner-time rejection.
535 JH/16 Fix logging of delivery remote port, when specified by router, under
538 PP/06 Repair manualroute's ability to take options in any order, even if one
539 is the name of a transport.
542 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
544 JH/17 Change the list-building routines interface to use the expanding-string
545 triplet model, for better allocation and copying behaviour.
547 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
548 Previously it was constructed the first time a possibly-matching string
549 was met in the configuration file input during startup; now it is done
552 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
553 compatible one, to avoid the (poorly documented) possibility of a config
554 file in the working directory redirecting the DB files, possibly correpting
555 some existing file. CVE-2017-10140 assigned for BDB.
557 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
558 cache-hot. Previously, although the result was properly cached, the
559 initial verify call returned a defer.
561 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
562 the main verify for receipient in uncached-mode.
564 JH/22 Retire historical build files to an "unsupported" subdir. These are
565 defined as "ones for which we have no current evidence of testing".
567 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
568 if present. Previously it was ignored.
570 JH/24 Start using specified-initialisers in C structure init coding. This is
571 a C99 feature (it's 2017, so now considered safe).
573 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
574 if was a fixed-sized field and bitmask ops via macros; it is now more
577 PP/07 GitHub PR 56: Apply MariaDB build fix.
578 Patch provided by Jaroslav Škarvada.
580 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
581 during Coverity cleanups [4.87 JH/47]
582 Diagnosis and fix provided by Michael Fischer v. Mollard.
584 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
585 the right size to place the terminating semicolon on its own folded
586 line, the header hash was calculated to an incorrect value thanks to
587 the (relaxed) space the fold became.
589 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
590 and confused the parent.
592 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
593 which could crash as a result. This could lead to undeliverable messages.
595 JH/28 Logging: "next input sent too soon" now shows where input was truncated
598 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
599 matters on fast-turnover and PID-randomising systems, which were getting
600 out-of-order delivery.
602 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
603 a possibly-overlapping copy. The symptom was that "Remote host closed
604 connection in response to HELO" was logged instead of the actual 4xx
607 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
608 Previously only that bufferd was discarded, resulting in SYMTP command
611 JH/32 DKIM: when a message has multiple signatures matching an identity given
612 in dkim_verify_signers, run the dkim acl once for each. Previously only
613 one run was done. Bug 2189.
615 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
616 "panic the current process" to "deliberately defer". The panic log is
617 still written with the problem list name; the mail and reject logs now
618 get a temp-reject line for the message that was being handled, saying
619 something like "domains check lookup or other defer". The SMTP 451
620 message is still "Temporary local problem".
622 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
623 A crafted sequence of BDAT commands could result in in-use memory beeing
624 freed. CVE-2017-16943.
626 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
627 from SMTP input. Previously it was always done; now only done for DATA
628 and not BDAT commands. CVE-2017-16944.
630 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
631 to the message (such as an overlong header line). Previously this was
632 not done and we did not exit BDAT mode. Followon from the previous item
633 though a different problem.
639 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
640 than -2003 did; needs libidn2 in addition to libidn.
642 JH/02 The path option on a pipe transport is now expanded before use.
644 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
645 Patch provided by "Björn", documentation fix added too.
647 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
648 missing a wire-to-host endian conversion.
650 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
651 close after a BDAT command line could be taken as a following command,
652 giving a synch failure. Fix by only checking for synch immediately
653 before acknowledging the chunk.
655 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
656 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
657 macro. Patches provided by Josh Soref.
659 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
660 Previously we did not; the RFC seems ambiguous and VRFY is not listed
661 by IANA as a service extension. However, John Klensin suggests that we
664 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
665 the dkim code may be unix-mode line endings rather than smtp wire-format
666 CRLF, so prepend a CR to any bare LF.
668 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
669 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
671 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
672 the first were themselves being wrongly included in the feed into dkim
673 processing; with most chunk sizes in use this resulted in an incorrect
674 body hash calculated value.
676 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
677 DKIM signature block, for verification. Although advised against by
678 standards it is specifically not ruled illegal.
680 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
682 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
683 missing a body hash (the bh= tag).
685 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
686 It seems that HAProxy sends the Proxy Protocol information in clear and
687 only then does a TLS startup, so do the same.
689 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
690 TCP connections (such as for Spamd) unless the daemon successfully set
691 Fast Open mode on its listening sockets. This fixes breakage seen on
692 too-old kernels or those not configured for Fast Open, at the cost of
693 requiring both directions being enabled for TFO, and TFO never being used
694 by non-daemon-related Exim processes.
696 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
697 endings, at least on the first header line. Try to canonify any that get
698 past that check, despite the cost.
700 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
701 now limited to an arbitrary five deep, while parsing addresses with the
702 strip_excess_angle_brackets option enabled.
704 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
705 instead leave the unprompted TLS handshake in socket buffer for the
706 TLS library to consume.
708 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
710 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
712 JH/16 Drop variables when they go out of scope. Memory management drops a whole
713 region in one operation, for speed, and this leaves assigned pointers
714 dangling. Add checks run only under the testsuite which checks all
715 variables at a store-reset and panics on a dangling pointer; add code
716 explicitly nulling out all the variables discovered. Fixes one known
717 bug: a transport crash, where a dangling pointer for $sending_ip_address
718 originally assigned in a verify callout, is re-used.
720 PP/06 Drop '.' from @INC in various Perl scripts.
722 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
724 PP/08 Reduce a number of compilation warnings under clang; building with
725 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
726 should be warning-free.
728 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
730 HS/01 Fix portability problems introduced by PP/08 for platforms where
731 realloc(NULL) is not equivalent to malloc() [SunOS et al].
733 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
734 chunk. This allows us to accept broken chunked messages. We need a more
735 general solution here.
737 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
738 already-broken messages in the queue.
740 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
742 JH/19 Fix reference counting bug in routing-generated-address tracking.
748 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
749 supports it and a size is available (ie. the sending peer gave us one).
751 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
752 years of being deprecated). The replacements are the ACLs
753 acl_smtp_mime and acl_not_smtp_mime.
755 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
756 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
757 or even an in-clear connection were permitted. Now, if the host lookup
758 was dnssec and dane was requested then the host is only used if the
759 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
760 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
761 if one fails this test.
762 This means that a poorly-configured remote DNS will make it incommunicado;
763 but it protects against a DNS-interception attack on it.
765 JH/04 Bug 1810: make continued-use of an open smtp transport connection
766 non-noisy when a race steals the message being considered.
768 JH/05 If main configuration option tls_certificate is unset, generate a
769 self-signed certificate for inbound TLS connections.
771 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
772 in rewrites and routers.
774 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
775 and logged a warning sing 4.83; now they are a configuration file error.
777 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
778 (lacking @domain). Apply the same qualification processing as RCPT.
780 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
782 JH/10 Support ${sha256:} applied to a string (as well as the previous
785 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
786 a cutthrough deliver is pending, as we always want to make a connection.
787 This also avoids re-routing the message when later placing the cutthrough
788 connection after a verify cache hit.
789 Do not update it with the verify result either.
791 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
792 when routing results in more than one destination address.
794 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
795 signing (which inhibits the cutthrough capability). Previously only
796 the presence of an option was tested; now an expansion evaluating as
797 empty is permissible (obviously it should depend only on data available
798 when the cutthrough connection is made).
800 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
801 the relevant preceding SMTP command did not note the pipelining mode.
803 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
804 Previously they were not counted.
806 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
807 as one having no matching records. Previously we deferred the message
808 that needed the lookup.
810 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
811 distinguished as "(=".
813 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
814 for missing MX records. Previously it only worked for missing A records.
816 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
818 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
819 after the data-go-ahead and data-ack. Patch from Jason Betts.
821 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
822 even for a "none" policy. Patch from Tony Meyer.
824 JH/22 Fix continued use of a connection for further deliveries. If a port was
825 specified by a router, it must also match for the delivery to be
828 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
829 When one of the recipients of a message was unsuitable for the connection
830 (has no matching addresses), we lost track of needing to mark it
831 deferred. As a result mail would be lost.
833 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
835 JH/25 Decoding ACL controls is now done using a binary search; the source code
836 takes up less space and should be simpler to maintain. Merge the ACL
837 condition decode tables also, with similar effect.
839 JH/26 Fix problem with one_time used on a redirect router which returned the
840 parent address unchanged. A retry would see the parent address marked as
841 delivered, so not attempt the (identical) child. As a result mail would
844 JH/27 Fix a possible security hole, wherein a process operating with the Exim
845 UID can gain a root shell. Credit to http://www.halfdog.net/ for
846 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
849 JH/28 Enable {spool,log} filesystem space and inode checks as default.
850 Main config options check_{log,spool}_{inodes,space} are now
851 100 inodes, 10MB unless set otherwise in the configuration.
853 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
854 Previously it only applied to the main-section connection policy
857 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
859 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
860 by me. Added RFC7919 DH primes as an alternative.
862 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
863 are not in the system include path.
865 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
866 GnuTLS, when a session startup failed (eg because the client disconnected)
867 Exim did stdio operations after fclose. This was exposed by a recent
868 change which nulled out the file handle after the fclose.
870 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
871 signed directly by the cert-signing cert, rather than an intermediate
872 OCSP-signing cert. This is the model used by LetsEncrypt.
874 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
876 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
877 an incoming connection.
879 HS/02 Bug 1802: Do not half-close the connection after sending a request
882 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
883 fallback to "prime256v1".
885 JH/34 SECURITY: Use proper copy of DATA command in error message.
886 Could leak key material. Remotely exploitable. CVE-2016-9963.
892 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
893 and 3.4.4 - once the server is enabled to respond to an OCSP request
894 it does even when not requested, resulting in a stapling non-aware
895 client dropping the TLS connection.
897 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
898 support variable-length bit vectors. No functional change.
900 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
901 The I= interface field on outgoing lines is now after the H= remote
902 host field, same as incoming lines. There is a separate
903 outgoing_interface log selector which allows you to disable the
906 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
907 If not running log_selector +smtp_connection the mainlog would be held
908 open indefinitely after a "too many connections" event, including to a
909 deleted file after a log rotate. Leave the per net connection logging
910 leaving it open for efficiency as that will be quickly detected by the
911 check on the next write.
913 HS/01 Bug 1671: Fix post transport crash.
914 Processing the wait-<transport> messages could crash the delivery
915 process if the message IDs didn't exist for some reason. When
916 using 'split_spool_directory=yes' the construction of the spool
917 file name failed already, exposing the same netto behaviour.
919 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
920 mime_regex ACL conditions.
922 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
923 to DSN fail messages (bounces): remote IP, remote greeting, remote response
924 to HELO, local diagnostic string.
926 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
927 log line to debug. Even when configured with a tls authenticator many
928 client connections are expected to not authenticate in this way, so
929 an authenticate fail is not an error.
931 HS/02 Add the Exim version string to the process info. This way exiwhat
932 gives some more detail about the running daemon.
934 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
935 matter for fast-change records such as DNSBLs.
937 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
938 retry record, even if constant. There may be multiple transports with
939 different interface settings and the retry behaviour needs to be kept
942 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
944 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
946 JH/10 Bug 840: fix log_defer_output option of pipe transport
948 JH/11 Bug 830: use same host for all RCPTS of a message, even under
949 hosts_randomize. This matters a lot when combined with mua_wrapper.
951 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
952 ${quote_pgsql:<string>} operator.
954 JH/13 Bug 1708: avoid misaligned access in cached lookup.
956 JH/14 Change header file name for freeradius-client. Relevant if compiling
957 with Radius support; from the Gentoo tree and checked under Fedora.
959 JH/15 Bug 1712: Introduce $prdr_requested flag variable
961 JH/16 Bug 1714: Permit an empty string as expansion result for transport
962 option transport_filter, meaning no filtering.
964 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
966 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
967 defaults to "*" (all hosts). The variable is now available when not built
968 with TLS, default unset, mainly to enable keeping the testsuite sane.
969 If a server certificate is not supplied (via tls_certificate) an error is
970 logged, and clients will find TLS connections fail on startup. Presumably
971 they will retry in-clear.
972 Packagers of Exim are strongly encouraged to create a server certificate
973 at installation time.
975 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
976 with the $config_file variable.
978 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
979 in transport context, after the attempt, and per-recipient. The latter type
980 is per host attempted. The event data is the error message, and the errno
981 information encodes the lookup type (A vs. MX) used for the (first) host,
982 and the trailing two digits of the smtp 4xx response.
984 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
985 to write to mainlog (or rejectlog, paniclog) in the window between file
986 creation and permissions/ownership being changed. Particularly affects
987 installations where exicyclog is run as root, rather than exim user;
988 result is that the running daemon panics and dies.
990 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
992 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
993 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
994 "pri" and "weight". Note that the previous implicit priority given by the
995 list order is no longer honoured.
997 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1000 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1001 by defining SUPPORT_SOCKS.
1003 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1004 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1005 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1006 variables are renamed to proxy_{local,external}_{address,port}.
1008 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1009 for a build by defining SUPPORT_I18N
1011 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1012 of the query string, and make ${quote_redis:} do that quoting.
1014 JH/29 Move Events support from Experimental to mainline, enabled by default
1015 and removable for a build by defining DISABLE_EVENT.
1017 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1019 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1020 cached by the daemon.
1022 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1023 by defining LOOKUP_REDIS. The libhiredis library is required.
1025 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1026 keys are given for lookup.
1028 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1029 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1030 only supported when built with TLS support. The PolarSSL SHA routines
1031 are still used when the TLS library is too old for convenient support.
1033 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1034 openssl_options), for security. OpenSSL forces this from version 1.1.0
1035 server-side so match that on older versions.
1037 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1038 allocation for $value could be released as the expansion processing
1039 concluded, but leaving the global pointer active for it.
1041 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1042 and to use the domains and local_parts ACL conditions.
1044 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1045 incorrectly not doubled on cutthrough transmission, hence seen as a
1046 body-termination at the receiving system - resulting in truncated mails.
1047 Commonly the sender saw a TCP-level error, and retransmitted the message
1048 via the normal store-and-forward channel. This could result in duplicates
1049 received - but deduplicating mailstores were liable to retain only the
1050 initial truncated version.
1052 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1054 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1056 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1057 we're in there, support oversigning also; bug 1309.
1059 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1061 HS/04 Add support for keep_environment and add_environment options.
1063 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1064 either intentional arithmetic overflow during PRNG, or testing config-
1067 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1068 delivery resulted in actual delivery. Cancel cutthrough before DATA
1071 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1072 rejects a recipient: pass the reject to the originator.
1074 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1075 Many were false-positives and ignorable, but it's worth fixing the
1078 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1079 for the new environment-manipulation done at startup. Move the routines
1080 from being local to tls.c to being global via the os.c file.
1082 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1083 an extract embedded as result-arg for a map, the first arg for extract
1084 is unavailable so we cannot tell if this is a numbered or keyed
1085 extraction. Accept either.
1091 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1094 JH/02 The smtp transport option "multi_domain" is now expanded.
1096 JH/03 The smtp transport now requests PRDR by default, if the server offers
1099 JH/04 Certificate name checking on server certificates, when exim is a client,
1100 is now done by default. The transport option tls_verify_cert_hostnames
1101 can be used to disable this per-host. The build option
1102 EXPERIMENTAL_CERTNAMES is withdrawn.
1104 JH/05 The value of the tls_verify_certificates smtp transport and main options
1105 default to the word "system" to access the system default CA bundle.
1106 For GnuTLS, only version 3.0.20 or later.
1108 JH/06 Verification of the server certificate for a TLS connection is now tried
1109 (but not required) by default. The verification status is now logged by
1110 default, for both outbound TLS and client-certificate supplying inbound
1113 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1116 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1117 Status Notification (bounce) messages are now MIME format per RFC 3464.
1118 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1119 under the control of the dsn_advertise_hosts option, and routers may
1120 have a dsn_lasthop option.
1122 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1123 default, modifiable by a malware= option. The list separator for
1124 the options can now be changed in the usual way. Bug 68.
1126 JH/10 The smtp_receive_timeout main option is now expanded before use.
1128 JH/11 The incoming_interface log option now also enables logging of the
1129 local interface on delivery outgoing connections.
1131 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1132 if the interface and destination host and port all match.
1134 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1137 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1138 Patch from Andrew Lewis.
1140 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1141 now supports optional time-restrictions, weighting, and priority
1142 modifiers per server. Patch originally by <rommer@active.by>.
1144 JH/16 The spamd_address main option now supports a mixed list of local
1145 and remote servers. Remote servers can be IPv6 addresses, and
1146 specify a port-range.
1148 JH/17 Bug 68: The spamd_address main option now supports an optional
1149 timeout value per server.
1151 JH/18 Bug 1581: Router and transport options headers_add/remove can
1152 now have the list separator specified.
1154 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1157 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1160 JH/21 Support for the A6 type of dns record is withdrawn.
1162 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1163 rather than the verbs used.
1165 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1166 from 255 to 1024 chars.
1168 JH/24 Verification callouts now attempt to use TLS by default.
1170 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1171 are generic router options now. The defaults didn't change.
1173 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1174 Original patch from Alexander Shikoff, worked over by JH.
1176 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1177 files if ZCAT_COMMAND is not executable.
1179 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1181 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1183 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1184 Normally benign, it bites when the pair was led to by a CNAME;
1185 modern usage is to not canonicalize the domain to a CNAME target
1186 (and we were inconsistent anyway for A-only vs AAAA+A).
1188 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1190 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1191 when evaluating $sender_host_dnssec.
1193 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1194 $sender_helo_dnssec variable.
1196 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1198 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1200 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1202 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1203 documented as working, but never had. Support all but $spam_report.
1205 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1206 added for tls authenticator.
1208 HS/03 Add perl_taintmode main config option
1214 TL/01 When running the test suite, the README says that variables such as
1215 no_msglog_check are global and can be placed anywhere in a specific
1216 test's script, however it was observed that placement needed to be near
1217 the beginning for it to behave that way. Changed the runtest perl
1218 script to read through the entire script once to detect and set these
1219 variables, reset to the beginning of the script, and then run through
1220 the script parsing/test process like normal.
1222 TL/02 The BSD's have an arc4random API. One of the functions to induce
1223 adding randomness was arc4random_stir(), but it has been removed in
1224 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1225 function when detected.
1227 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1228 cause callback expansion.
1230 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1231 syntax errors in an expansion can be treated as a string instead of
1232 logging or causing an error, due to the internal use of bool_lax
1233 instead of bool when processing it.
1235 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1236 server certificates when making smtp deliveries.
1238 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1240 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1242 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1244 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1245 Merged patch from Sebastian Wiedenroth.
1247 JH/05 Fix results-pipe from transport process. Several recipients, combined
1248 with certificate use, exposed issues where response data items split
1249 over buffer boundaries were not parsed properly. This eventually
1250 resulted in duplicates being sent. This issue only became common enough
1251 to notice due to the introduction of connection certificate information,
1252 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1254 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1255 size buffer was used, resulting in syntax errors when an expansion
1258 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1259 version 3.3.6 or later.
1261 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1262 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1263 both become "event_action", the variables become $event_name, $event_data
1264 and $event_defer_errno. There is a new variable $verify_mode, usable in
1265 routers, transports and related events. The tls:cert event is now also
1266 raised for inbound connections, if the main configuration event_action
1269 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1270 early OCSP support, but no stapling (appears to be less than 1.0.0).
1272 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1273 server certificate names available under the smtp transport option
1274 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1277 JH/10 Time-related extraction expansions from certificates now use the main
1278 option "timezone" setting for output formatting, and are consistent
1279 between OpenSSL and GnuTLS compilations. Bug 1541.
1281 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1282 encoded parameter in the incoming message. Bug 1558.
1284 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1285 include certificate info, eximon was claiming there were spoolfile
1288 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1290 JH/14 Log delivery-related information more consistently, using the sequence
1291 "H=<name> [<ip>]" wherever possible.
1293 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1294 are problematic for Debian distribution, omit them from the release
1297 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1299 JH/16 Fix string representation of time values on 64bit time_t architectures.
1302 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1303 output list separator was used.
1308 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1309 checkers that were complaining about end of non-void function with no
1312 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1313 This was a regression introduced in 4.83 by another bugfix.
1315 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1317 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1318 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1324 TF/01 Correctly close the server side of TLS when forking for delivery.
1326 When a message was received over SMTP with TLS, Exim failed to clear up
1327 the incoming connection properly after forking off the child process to
1328 deliver the message. In some situations the subsequent outgoing
1329 delivery connection happened to have the same fd number as the incoming
1330 connection previously had. Exim would try to use TLS and fail, logging
1331 a "Bad file descriptor" error.
1333 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1334 utilities have not been installed.
1336 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1337 temporary space as the ACL may create new global variables.
1339 TL/01 LDAP support uses per connection or global context settings, depending
1340 upon the detected version of the libraries at build time.
1342 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1343 to extract and use the src ip:port in logging and expansions as if it
1344 were a direct connection from the outside internet. PPv2 support was
1345 updated based on HAProxy spec change in May 2014.
1347 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1349 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1350 Properly escape header and check for NULL return.
1352 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1355 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1357 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1358 characters in header names, implemented as a verify condition.
1359 Contributed by Michael Fischer v. Mollard.
1361 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1362 results permerror and temperror. Previous values are deprecated but
1363 still accepted. In a future release, err_perm and err_temp will be
1364 completely removed, which will be a backward incompatibility if the
1365 ACL tests for either of these two old results. Patch contributed by
1366 user bes-internal on the mailing list.
1368 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1370 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1371 selectors, in both main and reject logs.
1373 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1376 JH/07 Add malware type "sock" for talking to simple daemon.
1378 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1380 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1381 routers/transports under cutthrough routing.
1383 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1384 numbers. Touch up "bool" conditional to keep the same definition.
1386 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1388 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1390 JH/12 Expand items in router/transport headers_add or headers_remove lists
1391 individually rather than the list as a whole. Bug 1452.
1393 Required for reasonable handling of multiple headers_ options when
1394 they may be empty; requires that headers_remove items with embedded
1395 colons must have them doubled (or the list-separator changed).
1397 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1398 view the policy declared in the DMARC record. Currently, $dmarc_status
1399 is a combined value of both the record presence and the result of the
1402 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1404 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1405 dnslookup router and the smtp transport (applying to the forward
1408 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1409 of ldap servers used for a specific lookup. Patch provided by Heiko
1412 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1413 New variable $lookup_dnssec_authenticated for observability.
1415 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1416 Patch submitted by Lars Timman.
1418 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1420 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1421 Requires trusted mode and valid format message id, aborts otherwise.
1422 Patch contributed by Heiko Schlichting.
1424 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1425 certextract with support for various fields. Bug 1358.
1427 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1428 is requested by default, modifiable by smtp transport option
1431 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1432 operate on certificate variables to give certificate fingerprints
1433 Also new ${sha256:cert_variable}.
1435 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1437 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1438 Christian Aistleitner.
1440 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1442 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1443 file. Patch from Wolfgang Breyha.
1445 JH/25 Expand the coverage of the delivery $host and $host_address to
1446 client authenticators run in verify callout. Bug 1476.
1448 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1449 align with daemon_smtp_ports. Bug 72.
1451 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1452 support and error reporting did not work properly.
1454 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1455 and is readable. Patch from Andrew Colin Kissa.
1457 TL/14 Enhance documentation of ${run expansion and how it parses the
1458 commandline after expansion, particularly in the case when an
1459 unquoted variable expansion results in an empty value.
1461 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1463 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1464 by renaming away from it. Addresses GH issue 15, reported by
1467 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1469 TL/15 SECURITY: prevent double expansion in math comparison functions
1470 (can expand unsanitized data). Not remotely exploitable.
1477 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1479 PP/02 Make -n do something, by making it not do something.
1480 When combined with -bP, the name of an option is not output.
1482 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1485 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1486 $sender_host_name and config options to manage this, and basic check
1489 PP/05 DSCP support for outbound connections and control modifier for inbound.
1491 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1492 (Only plugin which currently uses this is kerberos4, which nobody should
1493 be using, but we should make it available and other future plugins might
1494 conceivably use it, even though it would break NAT; stuff *should* be
1495 using channel bindings instead).
1497 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1498 name; added for Sendmail compatibility; requires admin caller.
1499 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1500 just ignore it); requires trusted caller.
1501 Also parse but ignore: -Ac -Am -X<logfile>
1504 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1506 TL/02 Add +smtp_confirmation as a default logging option.
1508 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1509 Patch by Magnus Holmgren from 2007-02-20.
1511 TL/04 Bugzilla 1281 - Spec typo.
1512 Bugzilla 1283 - Spec typo.
1513 Bugzilla 1290 - Spec grammar fixes.
1515 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1517 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1519 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1520 mailing list by by Dmitry Isaikin.
1522 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1524 JH/02 Support "G" suffix to numbers in ${if comparisons.
1526 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1528 NM/01 Bugzilla 1197 - Spec typo
1529 Bugzilla 1196 - Spec examples corrections
1531 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1533 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1534 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1537 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1538 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1540 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1541 "acl {{name}{arg}...}", and optional args on acl condition
1544 JH/05 Permit multiple router/transport headers_add/remove lines.
1546 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1548 JH/07 Avoid using a waiting database for a single-message-only transport.
1549 Performance patch from Paul Fisher. Bugzilla 1262.
1551 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1554 JH/09 Add $headers_added variable, with content from use of ACL modifier
1555 add_header (but not yet added to the message). Bugzilla 199.
1557 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1558 Pulled from Bugzilla 817 by Wolfgang Breyha.
1560 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1562 (nb: this is the same fix as in Exim 4.80.1)
1564 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1567 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1570 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1572 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1573 advertise SMTP AUTH mechanism to us, instead of a generic
1574 protocol violation error. Also, make Exim more robust to bad
1575 data from the Dovecot auth socket.
1577 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1579 When a queue runner is handling a message, Exim first routes the
1580 recipient addresses, during which it prunes them based on the retry
1581 hints database. After that it attempts to deliver the message to
1582 any remaining recipients. It then updates the hints database using
1585 So if a recipient address works intermittently, it can get repeatedly
1586 deferred at routing time. The retry hints record remains fresh so the
1587 address never reaches the final cutoff time.
1589 This is a fairly common occurrence when a user is bumping up against
1590 their storage quota. Exim had some logic in its local delivery code
1591 to deal with this. However it did not apply to per-recipient defers
1592 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1594 This change adds a proper retry rule check during routing so that the
1595 final cutoff time is checked against the message's age. We only do
1596 this check if there is an address retry record and there is not a
1597 domain retry record; this implies that previous attempts to handle
1598 the address had the retry_use_local_parts option turned on. We use
1599 this as an approximation for the destination being like a local
1600 delivery, as in LMTP.
1602 I suspect this new check makes the old local delivery cutoff check
1603 redundant, but I have not verified this so I left the code in place.
1605 TF/02 Correct gecos expansion when From: is a prefix of the username.
1607 Test 0254 submits a message to Exim with the header
1611 When I ran the test suite under the user fanf2, Exim expanded
1612 the header to contain my full name, whereas it should have added
1613 a Resent-Sender: header. It erroneously treats any prefix of the
1614 username as equal to the username.
1616 This change corrects that bug.
1618 GF/01 DCC debug and logging tidyup
1619 Error conditions log to paniclog rather than rejectlog.
1620 Debug lines prefixed by "DCC: " to remove any ambiguity.
1622 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1624 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1625 Bug spotted by Jeremy Harris; was flawed since initial commit.
1626 Would have resulted in OCSP responses post-SNI triggering an Exim
1627 NULL dereference and crash.
1629 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1631 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1632 Bug detection, analysis and fix by Samuel Thibault.
1633 Bugzilla 1331, Debian bug #698092.
1635 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1637 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1638 Server implementation by Todd Lyons, client by JH.
1639 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1640 config variable "prdr_enable" controls whether the server
1641 advertises the facility. If the client requests PRDR a new
1642 acl_data_smtp_prdr ACL is called once for each recipient, after
1643 the body content is received and before the acl_smtp_data ACL.
1644 The client is controlled by both of: a hosts_try_prdr option
1645 on the smtp transport, and the server advertisement.
1646 Default client logging of deliveries and rejections involving
1647 PRDR are flagged with the string "PRDR".
1649 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1650 fclose(). Diagnosis by Todd Lyons.
1652 PP/17 Update configure.default to handle IPv6 localhost better.
1653 Patch by Alain Williams (plus minor tweaks).
1656 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1657 This is now consistent with GnuTLS, and is now documented: the
1658 previous undocumented portable approach to treating the option as
1659 unset was to force an expansion failure. That still works, and
1660 an empty string is now equivalent.
1662 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1663 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1664 not performing validation itself.
1666 PP/20 Added force_command boolean option to pipe transport.
1667 Patch from Nick Koston, of cPanel Inc.
1669 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1672 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1674 PP/21 Fix eximon continuous updating with timestamped log-files.
1675 Broken in a format-string cleanup in 4.80, missed when I repaired the
1676 other false fix of the same issue.
1677 Report and fix from Heiko Schlichting.
1680 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1681 Report from Prashanth Katuri.
1683 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1684 It's SecureTransport, so affects any MacOS clients which use the
1685 system-integrated TLS libraries, including email clients.
1687 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1688 using a MIME ACL for non-SMTP local injection.
1689 Report and assistance in diagnosis by Warren Baker.
1691 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1693 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1695 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1696 last id that failed so it may be referenced in subsequent ACL's.
1698 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1701 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1702 ldap library initialization, allowing self-signed CA's to be
1703 used. Also properly sets require_cert option later in code by
1704 using NULL (global ldap config) instead of ldap handle (per
1705 session). Bug diagnosis and testing by alxgomz.
1707 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1708 the src/util/ subdirectory.
1710 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1711 renamed to Transport Post Delivery Action by Jeremy Harris, as
1714 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1715 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1716 redis_servers = needs to be configured which will be used by the redis
1717 lookup. Patch from Warren Baker, of The Packet Hub.
1719 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1721 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1722 hostname or reverse DNS when processing a host list. Used suggestions
1723 from multiple comments on this bug.
1725 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1727 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1728 Missed a few lines, added it to make the runtest require no keyboard
1731 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1732 contains upper case chars. Make router use caseful_local_part.
1734 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1735 support when GnuTLS has been built with p11-kit.
1741 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1743 This, or similar/improved, will also be change PP/11 of 4.82.
1749 PP/01 Handle short writes when writing local log-files.
1750 In practice, only affects FreeBSD (8 onwards).
1751 Bugzilla 1053, with thanks to Dmitry Isaikin.
1753 NM/01 Bugzilla 949 - Documentation tweak
1755 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1758 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1760 PP/02 Implemented gsasl authenticator.
1762 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1764 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1765 `pkg-config foo` for cflags/libs.
1767 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1768 with rest of GSASL and with heimdal_gssapi.
1770 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1771 `pkg-config foo` for cflags/libs for the TLS implementation.
1773 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1774 properties get this fed in as external SSF. A number of robustness
1775 and debugging improvements to the cyrus_sasl authenticator.
1777 PP/08 cyrus_sasl server now expands the server_realm option.
1779 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1780 Patch by Jeremy Harris.
1782 PP/10 Added dbmjz lookup type.
1784 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1786 PP/12 MAIL args handles TAB as well as SP, for better interop with
1787 non-compliant senders.
1788 Analysis and variant patch by Todd Lyons.
1790 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1791 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1792 Patch from Dirk Mueller <dmueller@suse.com>
1794 PP/13 tls_peerdn now print-escaped for spool files.
1795 Observed some $tls_peerdn in wild which contained \n, which resulted
1796 in spool file corruption.
1798 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1799 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1800 or write after TLS renegotiation, which otherwise led to messages
1803 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1804 as a tracking header (ie: a signed header comes before the signature).
1805 Patch from Wolfgang Breyha.
1807 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1808 comma-sep list; embedded commas doubled.
1810 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1812 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1814 Report and patch from Dmitry Banschikov.
1816 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1817 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1818 failures. We appear to now support TLS1.1+ with Exim.
1820 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1821 lets Exim select keys and certificates based upon TLS SNI from client.
1822 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1823 before an outbound SMTP session. New log_selector, +tls_sni.
1825 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1826 NULL dereference. Report and patch from Alun Jones.
1828 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1829 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1830 Not seeing resolver debug output on NetBSD, but suspect this is a
1831 resolver implementation change.
1833 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1834 Left warnings. Added "eximon gdb" invocation mode.
1836 PP/21 Defaulting "accept_8bitmime" to true, not false.
1838 PP/22 Added -bw for inetd wait mode support.
1840 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1841 locate the relevant includes and libraries. Made this the default.
1843 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1844 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1846 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1847 This may cause build issues on older platforms.
1849 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1850 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1851 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1852 Added SNI support via GnuTLS too.
1853 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1855 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1857 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1858 Applied second patch from Janne, implementing suggestion to default
1859 multiple-strings-in-record handling to match SPF spec.
1861 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1863 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1864 read-only, out of scope).
1865 Patch from Wolfgang Breyha, report from Stuart Northfield.
1867 PP/29 Fix three issues highlighted by clang analyser static analysis.
1868 Only crash-plausible issue would require the Cambridge-specific
1869 iplookup router and a misconfiguration.
1870 Report from Marcin Mirosław.
1872 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1874 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1875 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1876 As part of this, removing so much warning spew let me fix some minor
1877 real issues in debug logging.
1879 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1880 assignment on my part. Fixed.
1882 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1883 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1884 Janne Snabb (who went above and beyond: thank you).
1886 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1887 string otherwise requires a connection and a bunch more work and it's
1888 relatively easy to get wrong. Should also expose TLS library linkage
1891 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1892 64-bit ${eval} (JH/03).
1894 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1895 GNU libc to support some of the 64-bit stuff, should not lead to
1896 conflicts. Defined before os.h is pulled in, so if a given platform
1897 needs to override this, it can.
1899 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1900 protection layer was required, which is not implemented.
1901 Bugzilla 1254, patch from Wolfgang Breyha.
1903 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1904 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1905 tls_dhparam take prime identifiers. Also unbreak combination of
1906 OpenSSL+DH_params+TLSSNI.
1908 PP/39 Disable SSLv2 by default in OpenSSL support.
1914 PP/01 Solaris build fix for Oracle's LDAP libraries.
1915 Bugzilla 1109, patch from Stephen Usher.
1917 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1919 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1922 TF/02 Fix a couple more cases where we did not log the error message
1923 when unlink() failed. See also change 4.74-TF/03.
1925 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1926 lock up or crash if it happened to be inside a call to libc when it
1927 got a SIGUSR1 from exiwhat.
1929 The SIGUSR1 handler appends the current process status to the process
1930 log which is later printed by exiwhat. It used to use the general
1931 purpose logging code to do this, but several functions it calls are
1932 not safe for signals.
1934 The new output code in the SIGUSR1 handler is specific to the process
1935 log, and simple enough that it's easy to inspect for signal safety.
1936 Removing some special cases also simplifies the general logging code.
1937 Removing the spurious timestamps from the process log simplifies
1940 TF/04 Improved ratelimit ACL condition.
1942 The /noupdate option has been deprecated in favour of /readonly which
1943 has clearer semantics. The /leaky, /strict, and /readonly update modes
1944 are mutually exclusive. The update mode is no longer included in the
1945 database key; it just determines when the database is updated. (This
1946 means that when you upgrade Exim will forget old rate measurements.)
1948 Exim now checks that the per_* options are used with an update mode that
1949 makes sense for the current ACL. For example, when Exim is processing a
1950 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1951 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1952 must specify per_mail/readonly. If you omit the update mode it defaults to
1953 /leaky where that makes sense (as before) or /readonly where required.
1955 The /noupdate option is now undocumented but still supported for
1956 backwards compatibility. It is equivalent to /readonly except that in
1957 ACLs where /readonly is required you may specify /leaky/noupdate or
1958 /strict/noupdate which are treated the same as /readonly.
1960 A useful new feature is the /count= option. This is a generalization
1961 of the per_byte option, so that you can measure the throughput of other
1962 aggregate values. For example, the per_byte option is now equivalent
1963 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1965 The per_rcpt option has been generalized using the /count= mechanism
1966 (though it's more complicated than the per_byte equivalence). When it is
1967 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1968 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1969 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1970 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1971 Note that using per_rcpt with a non-readonly update mode in more than
1972 one ACL will cause the recipients to be double-counted. (The per_mail
1973 and per_byte options don't have this problem.)
1975 The handling of very low rates has changed slightly. If the computed rate
1976 is less than the event's count (usually one) then this event is the first
1977 after a long gap. In this case the rate is set to the same as this event's
1978 count, so that the first message of a spam run is counted properly.
1980 The major new feature is a mechanism for counting the rate of unique
1981 events. The new per_addr option counts the number of different
1982 recipients that someone has sent messages to in the last time period. It
1983 behaves like per_rcpt if all the recipient addresses are different, but
1984 duplicate recipient addresses do not increase the measured rate. Like
1985 the /count= option this is a general mechanism, so the per_addr option
1986 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1987 example, measure the rate that a client uses different sender addresses
1988 with the options per_mail/unique=$sender_address. There are further
1989 details in the main documentation.
1991 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1993 TF/06 Removed a few PCRE remnants.
1995 TF/07 Automatically extract Exim's version number from tags in the git
1996 repository when doing development or release builds.
1998 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1999 Bugzilla 879. Patch from Paul Fisher.
2001 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2002 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2005 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2007 PP/05 Variable $av_failed, true if the AV scanner deferred.
2008 Bugzilla 1078. Patch from John Horne.
2010 PP/06 Stop make process more reliably on build failure.
2011 Bugzilla 1087. Patch from Heiko Schlittermann.
2013 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2014 Bugzilla 1089. Patch from Heiko Schlittermann.
2016 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2017 Bugzilla 1131. Patch from Holger Weiß.
2019 PP/09 Handle IPv6 addresses with SPF.
2020 Bugzilla 860. Patch from Wolfgang Breyha.
2022 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2024 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2027 PP/11 match_* no longer expand right-hand-side by default.
2028 New compile-time build option, EXPAND_LISTMATCH_RHS.
2029 New expansion conditions, "inlist", "inlisti".
2031 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2033 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2035 PP/14 fix log_write() format string regression from TF/03.
2036 Bugzilla 1152. Patch from Dmitry Isaikin.
2042 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2044 PP/02 Harmonised TLS library version reporting; only show if debugging.
2045 Layout now matches that introduced for other libraries in 4.74 PP/03.
2047 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2049 PP/04 New "dns_use_edns0" global option.
2051 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2054 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2055 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2057 TK/01 Updated PolarSSL code to 0.14.2.
2058 Bugzilla 1097. Patch from Andreas Metzler.
2060 PP/07 Catch divide-by-zero in ${eval:...}.
2061 Fixes bugzilla 1102.
2063 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2066 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2067 format-string attack -- SECURITY: remote arbitrary code execution.
2069 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2070 time unintentionally subject to list matching rules, letting the header
2071 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2072 arbitrary string expansion). This allowed for information disclosure.
2074 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2075 INT_MIN/-1 -- value coerced to INT_MAX.
2081 NM/01 Workaround for PCRE version dependency in version reporting
2084 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2085 This fixes portability to compilers other than gcc, notably
2086 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2088 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2089 makefiles for portability to HP-UX and POSIX correctness.
2091 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2092 Also via indented variable definition in the Makefile.
2093 (Debugging by Oliver Heesakkers).
2095 PP/02 Restore caching of spamd results with expanded spamd_address.
2096 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2098 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2099 Improves build reliability. Fix from: Frank Elsner
2101 NM/02 Fix wide character breakage in the rfc2047 coding
2102 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2104 NM/03 Allow underscore in dnslist lookups
2105 Fixes bug 1026. Patch from Graeme Fowler
2107 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2108 Code patches from Adam Ciarcinski of NetBSD.
2110 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2113 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2114 is logged, to avoid truncation. Patch from John Horne.
2116 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2117 Patch from Jakob Hirsch.
2119 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2120 SQL string expansion failure details.
2121 Patch from Andrey Oktyabrski.
2123 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2124 Patch from Simon Arlott.
2126 PP/09 New lookups functionality failed to compile on old gcc which rejects
2127 extern declarations in function scope.
2128 Patch from Oliver Fleischmann
2130 PP/10 Use sig_atomic_t for flags set from signal handlers.
2131 Check getgroups() return and improve debugging.
2132 Fixed developed for diagnosis in bug 927 (which turned out to be
2135 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2136 Patch from Mark Zealey.
2138 PP/12 Bugzilla 1056: Improved spamd server selection.
2139 Patch from Mark Zealey.
2141 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2142 Based on patch from Heiko Schlittermann.
2144 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2145 Patch from Uwe Doering, sign-off by Michael Haardt.
2147 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2148 variable declaration deep within a block. Bug and patch from
2151 PP/15 lookups-Makefile IRIX compatibility coercion.
2153 PP/16 Make DISABLE_DKIM build knob functional.
2155 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2156 Patch by Simon Arlott
2158 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2159 variable argument macros. Our copy now differs from upstream.
2165 TF/01 Failure to get a lock on a hints database can have serious
2166 consequences so log it to the panic log.
2168 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2169 controlled using the smtp_confirmation log selector.
2171 TF/03 Include the error message when we fail to unlink a spool file.
2173 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2174 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2175 for maintaining out-of-tree patches for some time.
2177 PP/01 Bugzilla 139: Documentation and portability issues.
2178 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2179 Handle per-OS dynamic-module compilation flags.
2181 PP/02 Let /dev/null have normal permissions.
2182 The 4.73 fixes were a little too stringent and complained about the
2183 permissions on /dev/null. Exempt it from some checks.
2184 Reported by Andreas M. Kirchwitz.
2186 PP/03 Report version information for many libraries, including
2187 Exim version information for dynamically loaded libraries. Created
2188 version.h, now support a version extension string for distributors
2189 who patch heavily. Dynamic module ABI change.
2191 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2192 privilege escalation vulnerability whereby the Exim run-time user
2193 can cause root to append content of the attacker's choosing to
2196 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2199 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2200 If dropping privileges for untrusted macros, we disabled normal logging
2201 on the basis that it would fail; for the Exim run-time user, this is not
2202 the case, and it resulted in successful deliveries going unlogged.
2203 Fixed. Reported by Andreas Metzler.
2209 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2210 only prepend for the Resent-* case. Fixes regression introduced in
2211 Exim 4.70 by NM/22 for Bugzilla 607.
2213 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2214 increasing numbers of administrators be bitten by this.
2216 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2218 PP/03 Bugzilla 994: added openssl_options main configuration option.
2220 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2222 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2224 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2226 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2227 without a peer certificate, leading to a segfault because of an
2228 assumption that peers always have certificates. Be a little more
2229 paranoid. Problem reported by Martin Tscholak.
2231 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2232 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2233 NB: ClamAV planning to remove STREAM in "middle of 2010".
2234 CL also introduces -bmalware, various -d+acl logging additions and
2235 more caution in buffer sizes.
2237 PP/09 Implemented reverse_ip expansion operator.
2239 PP/10 Bugzilla 937: provide a "debug" ACL control.
2241 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2243 PP/12 Bugzilla 973: Implement --version.
2245 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2247 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2249 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2251 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2252 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2253 ignore trailing whitespace.
2255 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2257 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2260 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2261 Notification from Dr Andrew Aitchison.
2263 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2264 ExtendedDetectionInfo response format.
2265 Notification from John Horne.
2267 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2270 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2271 XSL and documented dependency on system catalogs, with examples of how
2274 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2277 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2278 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2279 configuration file which is writeable by the Exim user or group.
2281 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2282 of configuration files to cover files specified with the -C option if
2283 they are going to be used with root privileges, not just the default
2286 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2287 option (effectively making it always true).
2289 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2290 files to be used while preserving root privileges.
2292 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2293 that rogue child processes cannot use them.
2295 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2296 run-time user, instead of root.
2298 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2299 Exim run-time user without dropping privileges.
2301 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2302 result string, instead of calling string_vformat() twice with the same
2305 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2306 for other users. Others should always drop root privileges if they use
2307 -C on the command line, even for a whitelisted configure file.
2309 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2311 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2317 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2318 $data_path, and $header_path variables; fixed documentation bugs and
2321 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2322 exipick to access non-standard spools, including the "frozen" queue
2325 NM/01 Bugzilla 965: Support mysql stored procedures.
2326 Patch from Alain Williams
2328 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2330 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2331 Patch from Andreas Metzler
2333 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2334 Patch from Kirill Miazine
2336 NM/05 Bugzilla 671: Added umask to procmail example.
2338 JJ/03 installed exipick 20100323.0, fixing doc bug
2340 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2341 directory. Notification and patch from Dan Rosenberg.
2343 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2345 TK/02 Improve log output when DKIM signing operation fails.
2347 MH/01 Treat the transport option dkim_domain as a colon separated
2348 list, not as a single string, and sign the message with each element,
2349 omitting multiple occurences of the same signer.
2351 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2352 Bugzilla 985, 986. Patch by Simon Arlott
2354 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2355 Patch by Simon Arlott
2357 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2358 MBX locking. Notification from Dan Rosenberg.
2364 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2366 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2368 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2370 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2372 NM/04 Bugzilla 915: Flag broken perl installation during build.
2378 TK/01 Added patch by Johannes Berg that expands the main option
2379 "spamd_address" if it starts with a dollar sign.
2381 TK/02 Write list of recipients to X-Envelope-Sender header when building
2382 the mbox-format spool file for content scanning (suggested by Jakob
2385 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2386 (http://www.dcc-servers.net/) support via dccifd. Activated by
2387 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2389 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2390 by Mark Daniel Reidel <mr@df.eu>.
2392 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2393 When building exim an external PCRE library is now needed -
2394 PCRE is a system library on the majority of modern systems.
2395 See entry on PCRE_LIBS in EDITME file.
2397 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2398 conversation. Added nologin parameter to request.
2399 Patch contributed by Kirill Miazine.
2401 TF/01 Do not log submission mode rewrites if they do not change the address.
2403 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2405 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2406 log files in place. Contributed by Roberto Lima.
2408 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2410 TF/03 Bugzilla 615: When checking the local_parts router precondition
2411 after a local_part_suffix or local_part_prefix option, Exim now
2412 does not use the address's named list lookup cache, since this
2413 contains cached lookups for the whole local part.
2415 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2416 Robert Millan. Documentation is in experimental-spec.txt.
2418 TF/04 Bugzilla 668: Fix parallel build (make -j).
2420 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2422 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2423 Patch provided by Jan Srzednicki.
2425 TF/05 Leading white space used to be stripped from $spam_report which
2426 wrecked the formatting. Now it is preserved.
2428 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2429 that they are available at delivery time.
2431 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2433 TF/08 TLS error reporting now respects the incoming_interface and
2434 incoming_port log selectors.
2436 TF/09 Produce a more useful error message if an SMTP transport's hosts
2437 setting expands to an empty string.
2439 NM/06 Bugzilla 744: EXPN did not work under TLS.
2440 Patch provided by Phil Pennock.
2442 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2443 Patch provided by Richard Godbee.
2445 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2446 acl_smtp_notquit, added index entry.
2448 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2449 Patch provided by Eugene Bujak.
2451 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2452 accept(). Patch provided by Maxim Dounin.
2454 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2455 Patch provided by Phil Pennock.
2457 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2459 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2460 Patch provided by Brad "anomie" Jorsch.
2462 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2463 Patch provided by Dean Brooks.
2465 TK/05 Add native DKIM support (does not depend on external libraries).
2467 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2468 Patch provided by Graeme Fowler.
2470 NM/16 Bugzilla 851: Documentation example syntax fix.
2472 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2474 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2477 NM/19 Bugzilla 745: TLS version reporting.
2478 Patch provided by Phil Pennock.
2480 NM/20 Bugzilla 167: bool: condition support.
2481 Patch provided by Phil Pennock.
2483 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2484 clients. Patch provided by Phil Pennock.
2486 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2487 Patch provided by Brad "anomie" Jorsch.
2489 NM/23 Bugzilla 687: Fix misparses in eximstats.
2490 Patch provided by Heiko Schlittermann.
2492 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2493 Patch provided by Heiko Schlittermann.
2495 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2496 plus update to original patch.
2498 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2500 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2501 Patch provided by David Brownlee.
2503 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2505 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2507 NM/30 Bugzilla 888: TLS documentation bugfixes.
2509 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2511 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2512 Unlike the original bugzilla I have changed all shell scripts in src tree.
2514 NM/33 Bugzilla 898: Transport filter timeout fix.
2515 Patch by Todd Rinaldo.
2517 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2518 Patch by Serge Demonchaux.
2520 NM/35 Bugzilla 39: Base64 decode bug fixes.
2521 Patch by Jakob Hirsch.
2523 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2525 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2527 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2529 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2535 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2536 ALT-N's libdkim that I have put here:
2537 http://duncanthrax.net/exim-experimental/
2539 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2540 were called 'true' and it seems that C99 defines that as a reserved
2541 keyword to be used with 'bool' variable types. That means you could
2542 not include C99-style headers which use bools without triggering
2543 build errors in sieve.c.
2545 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2546 as mailq or other aliases. Changed the --help handling significantly
2547 to do whats expected. exim_usage() emits usage/help information.
2549 SC/01 Added the -bylocaldomain option to eximstats.
2551 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2553 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2555 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2561 PH/01 Another patch from the Sieve maintainer.
2563 PH/02 When an IPv6 address is converted to a string for single-key lookup
2564 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2565 dots are used instead of colons so that keys in lsearch files need not
2566 contain colons. This was done some time before quoting was made available
2567 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2568 (notated using the quote facility) so as to distinguish them from IPv4
2569 keys. This meant that lookups for IP addresses in host lists did not work
2570 for iplsearch lookups.
2572 This has been fixed by arranging for IPv6 addresses to be expressed with
2573 colons if the lookup type is iplsearch. This is not incompatible, because
2574 previously such lookups could never work.
2576 The situation is now rather anomalous, since one *can* have colons in
2577 ordinary lsearch keys. However, making the change in all cases is
2578 incompatible and would probably break a number of configurations.
2580 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2583 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2584 conversion specification without a maximum field width, thereby enabling
2585 a rogue spamd server to cause a buffer overflow. While nobody in their
2586 right mind would setup Exim to query an untrusted spamd server, an
2587 attacker that gains access to a server running spamd could potentially
2588 exploit this vulnerability to run arbitrary code as the Exim user.
2590 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2591 $primary_hostname instead of what libspf2 thinks the hosts name is.
2593 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2594 a directory entry by the name of the lookup key. Previously, if a
2595 symlink pointed to a non-existing file or a file in a directory that
2596 Exim lacked permissions to read, a lookup for a key matching that
2597 symlink would fail. Now it is enough that a matching directory entry
2598 exists, symlink or not. (Bugzilla 503.)
2600 PH/03 The body_linecount and body_zerocount variables are now exported in the
2603 PH/04 Added the $dnslist_matched variable.
2605 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2606 This means they are set thereafter only if the connection becomes
2609 PH/06 Added the client_condition to authenticators so that some can be skipped
2610 by clients under certain conditions.
2612 PH/07 The error message for a badly-placed control=no_multiline_responses left
2613 "_responses" off the end of the name.
2615 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2617 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2618 (without spaces) instead of just copying the configuration text.
2620 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2622 PH/11 Added $max_received_linelength.
2624 PH/12 Added +ignore_defer and +include_defer to host lists.
2626 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2627 way in which PCRE > 7.0 is built.
2629 PH/14 Implemented queue_only_load_latch.
2631 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2632 MAIL command. The effect was to mangle the value on 64-bit systems.
2634 PH/16 Another patch from the Sieve maintainer.
2636 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2638 PH/18 If a system quota error occurred while trying to create the file for
2639 a maildir delivery, the message "Mailbox is full" was not appended to the
2640 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2641 only to a quota excession during the actual writing of the file.
2643 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2644 characters?) which causes problems in log lines. The DN values are now
2645 passed through string_printing() before being added to log lines.
2647 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2648 and InterBase are left for another time.)
2650 PH/21 Added message_body_newlines option.
2652 PH/22 Guard against possible overflow in moan_check_errorcopy().
2654 PH/23 POSIX allows open() to be a macro; guard against that.
2656 PH/24 If the recipient of an error message contained an @ in the local part
2657 (suitably quoted, of course), incorrect values were put in $domain and
2658 $local_part during the evaluation of errors_copy.
2664 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2665 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2668 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2669 issue a MAIL command.
2671 PH/02 In an ACL statement such as
2673 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2675 if a client was not listed at all, or was listed with a value other than
2676 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2677 the condition was not true (as it should be), so access was not denied.
2678 The bug was that the ! inversion was incorrectly passed on to the second
2679 item. This has been fixed.
2681 PH/03 Added additional dnslists conditions == and =& which are different from
2682 = and & when the dns lookup returns more than one IP address.
2684 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2685 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2687 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2688 FSYNC, which compiles an option called disable_fsync that allows for
2689 bypassing fsync(). The documentation is heavily laced with warnings.
2691 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2693 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2694 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2695 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2696 including adding "make clean"; (3) Added -fPIC when compiling the test
2697 dynamically loaded module, to get rid of a warning.
2699 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2700 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2701 = 0s. The bug is otherwise harmless.
2703 PH/07 There was a bug in the dovecot authenticator such that the value of
2704 $auth1 could be overwritten, and so not correctly preserved, after a
2705 successful authentication. This usually meant that the value preserved by
2706 the server_setid option was incorrect.
2708 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2710 PH/09 Installed PCRE release 7.0.
2712 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2713 run for batched SMTP input. It is now run at the start of every message
2714 in the batch. While fixing this I discovered that the process information
2715 (output by running exiwhat) was not always getting set for -bs and -bS
2716 input. This is fixed, and it now also says "batched" for BSMTP.
2718 PH/11 Added control=no_pipelining.
2720 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2721 patch, slightly modified), and move the expansion of helo_data till after
2722 the connection is made in the smtp transport (so it can use these
2725 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2727 PH/14 Added log_selector = +pid.
2729 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2731 PH/16 Add ${if forany and ${if forall.
2733 PH/17 Added dsn_from option to vary the From: line in DSNs.
2735 PH/18 Flush SMTP output before performing a callout, unless control =
2736 no_callout_flush is set.
2738 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2739 was true (the default) a successful delivery failed to delete the retry
2740 item, thus causing premature timeout of the address. The bug is now
2743 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2745 PH/21 Long custom messages for fakedefer and fakereject are now split up
2746 into multiline responses in the same way that messages for "deny" and
2747 other ACL rejections are.
2749 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2750 with slight modification.
2752 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2753 draft, changing the syntax and factoring some duplicate code".
2755 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2756 for deliveries of the second and subsequent messages over the same SMTP
2759 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2760 ${reduce, with only minor "tidies".
2762 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2764 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2765 expansion side effects.
2767 PH/27 When a message times out after an over-quota error from an Exim-imposed
2768 quota, the bounce message says "mailbox is full". This message was not
2769 being given when it was a system quota that was exceeded. It now should
2772 MH/03 Made $recipients available in local_scan(). local_scan() already has
2773 better access to the recipient list through recipients_list[], but
2774 $recipients can be useful in postmaster-provided expansion strings.
2776 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2777 in the case of a MAIL command with additional options following the
2778 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2779 were accidentally chopped off.
2781 PH/29 SMTP synchronization checks are implemented when a command is read -
2782 there is a check that no more input is waiting when there shouldn't be
2783 any. However, for some commands, a delay in an ACL can mean that it is
2784 some time before the response is written. In this time, more input might
2785 arrive, invalidly. So now there are extra checks after an ACL has run for
2786 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2787 pipelining has not been advertised.
2789 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2791 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2792 returned in the SMTP response when smtp_return_error_details was false.
2793 This has been fixed.
2795 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2796 instead of the C I/O that was originally supplied, because problems were
2797 reported on Solaris.
2799 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2800 Exim which did not show up earlier: it was assuming that a call to
2801 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2802 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2803 was a macro that became an assignment, so it seemed to work. This has
2804 changed to a proper function call with a void return, hence the compile
2805 error. Exim's code has been fixed.
2807 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2810 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2812 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2814 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2815 because it turns out to be too big or there is a timeout) there is no
2816 "Completed" line in the log. When some messages of this type were
2817 selected by exigrep, they were listed as "not completed". Others were
2818 picked up by some special patterns. I have improved the selection
2819 criteria to be more general.
2821 PH/38 The host_find_failed option in the manualroute router can now be set
2822 to "ignore", to completely ignore a host whose IP address cannot be
2823 found. If all hosts are ignored, the behaviour is controlled by the new
2824 host_all_ignored option.
2826 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2827 homing or because of multiple MX records with /mx) generated more than
2828 one IP address, and the following item turned out to be the local host,
2829 all the secondary addresses of the first item were incorrectly removed
2830 from the list, along with the local host and any following hosts (which
2831 is what is supposed to happen).
2833 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2834 whoever called Exim into the -H file. In the case of the daemon it was
2835 behaving confusingly. When first started, it used values for whoever
2836 started the daemon, but after a SIGHUP it used the Exim user (because it
2837 calls itself on a restart). I have changed the code so that it now always
2840 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2841 message are rejected with the same error (e.g. no authentication or bad
2842 sender address), and a DATA command is nevertheless sent (as can happen
2843 with PIPELINING or a stupid MUA), the error message that was given to the
2844 RCPT commands is included in the rejection of the DATA command. This is
2845 intended to be helpful for MUAs that show only the final error to their
2848 PH/42 Another patch from the Sieve maintainer.
2850 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2851 Eximstats - Fixed some broken HTML links and added missing column headers
2853 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2854 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2856 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2858 PH/43 Yet another patch from the Sieve maintainer.
2860 PH/44 I found a way to check for a TCP/IP connection going away before sending
2861 the response to the final '.' that terminates a message, but only in the
2862 case where the client has not sent further data following the '.'
2863 (unfortunately, this is allowed). However, in many cases there won't be
2864 any further data because there won't be any more messages to send. A call
2865 to select() can be used: if it shows that the input is "ready", there is
2866 either input waiting, or the socket has been closed. An attempt to read
2867 the next input character can distinguish the two cases. Previously, Exim
2868 would have sent an OK response which the client would never have see.
2869 This could lead to message repetition. This fix should cure that, at
2870 least in a lot of common cases.
2872 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2873 advertised in response to EHLO.
2879 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2880 fixed by 4.65/MH/01 (is this a record?) are fixed:
2882 (i) An empty string was always treated as zero by the numeric comparison
2883 operators. This behaviour has been restored.
2885 (ii) It is documented that the numeric comparison operators always treat
2886 their arguments as decimal numbers. This was broken in that numbers
2887 starting with 0 were being interpreted as octal.
2889 While fixing these problems I realized that there was another issue that
2890 hadn't been noticed. Values of message_size_limit (both the global option
2891 and the transport option) were treated as octal if they started with 0.
2892 The documentation was vague. These values are now always treated as
2893 decimal, and I will make that clear in the documentation.
2899 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2900 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2903 MH/01 Don't check that the operands of numeric comparison operators are
2904 integers when their expansion is in "skipping" mode (fixes bug
2905 introduced by 4.64-PH/07).
2907 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2908 child addresses, Exim now panics and dies. Previously, because the count
2909 is held in a short int, deliveries were likely to be lost. As such a
2910 large number of recipients for a single message is ridiculous
2911 (performance will be very, very poor), I have chosen to impose a limit
2912 rather than extend the field.
2918 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2919 leftover -K file (the existence of which was triggered by #402).
2920 While we were at it, introduced process PID as part of the -K
2921 filename. This should rule out race conditions when creating
2924 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2925 processing considerably. Previous code took too long for large mails,
2926 triggering a timeout which in turn triggers #401.
2928 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2929 in the DK code in transports.c. sendfile() is not really portable,
2930 hence the _LINUX specificness.
2932 TF/01 In the add_headers option to the mail command in an Exim filter,
2933 there was a bug that Exim would claim a syntax error in any
2934 header after the first one which had an odd number of characters
2937 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2938 callout verification, Exim cached a "reject" for the entire domain. This
2939 is correct for most verifications, but it is not correct for a recipient
2940 verification with use_sender or use_postmaster set, because in that case
2941 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2942 case of MAIL FROM:<> rejection from other early rejections (e.g.
2943 rejection of HELO). When verifying a recipient using a non-null MAIL
2944 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2945 Whatever the result of the callout, the value of the domain cache is
2946 left unchanged (for any other kind of callout, getting as far as trying
2947 RCPT means that the domain itself is ok).
2949 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2952 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2953 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2954 session, the daemon ignores ECONNECT errors and logs others; it now
2955 ignores EPIPE as well.
2957 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2958 (quoted-printable decoding).
2960 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2961 later the small subsequent patch to fix an introduced bug.
2963 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2965 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2967 PH/08 An error is now given if message_size_limit is specified negative.
2969 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2970 to be given (somewhat) arbitrary names.
2972 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2975 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2976 miscellaneous code fixes
2978 PH/10 Added the log_reject_target ACL modifier to specify where to log
2981 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2982 hostname. This is wrong, because it relates to the incoming message (and
2983 probably the interface on which it is arriving) and not to the outgoing
2984 callout (which could be using a different interface). This has been
2985 changed to use the value of the helo_data option from the smtp transport
2986 instead - this is what is used when a message is actually being sent. If
2987 there is no remote transport (possible with a router that sets up host
2988 addresses), $smtp_active_hostname is used.
2990 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2991 tweaks were necessary in order to get it to work (see also 21 below):
2992 (a) The code assumed that strncpy() returns a negative number on buffer
2993 overflow, which isn't the case. Replaced with Exim's string_format()
2995 (b) There were several signed/unsigned issues. I just did the minimum
2996 hacking in of casts. There is scope for a larger refactoring.
2997 (c) The code used strcasecmp() which is not a standard C function.
2998 Replaced with Exim's strcmpic() function.
2999 (d) The code set only $1; it now sets $auth1 as well.
3000 (e) A simple test gave the error "authentication client didn't specify
3001 service in request". It would seem that Dovecot has changed its
3002 interface. Fortunately there's a specification; I followed it and
3003 changed what the client sends and it appears to be working now.
3005 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3008 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3009 address A is aliased to B and C, where B exists and C does not. Without
3010 -v the output is "A verified" because verification stops after a
3011 successful redirection if more than one address is generated. However,
3012 with -v the child addresses are also verified. Exim was outputting "A
3013 failed to verify" and then showing the successful verification for C,
3014 with its parentage. It now outputs "B failed to verify", showing B's
3015 parentage before showing the successful verification of C.
3017 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3018 look up a TXT record in a specific list after matching in a combined
3021 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3022 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3023 they consult the DNS. I had assumed they would set it the way they
3024 wanted; and indeed my experiments on Linux seem to show that in some
3025 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3026 To be on the safe side, however, I have now made the interface to
3027 host_find_byname() similar to host_find_bydns(), with an argument
3028 containing the DNS resolver options. The host_find_byname() function now
3029 sets these options at its start, just as host_find_bydns() does. The smtp
3030 transport options dns_qualify_single and dns_search_parents are passed to
3031 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3032 of host_find_byname() use the default settings of RES_DEFNAMES
3033 (qualify_single) but not RES_DNSRCH (search_parents).
3035 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3036 spool_read_header() do less string testing, by means of a preliminary
3037 switch on the second character of optional "-foo" lines. (This is
3038 overdue, caused by the large number of possibilities that now exist.
3039 Originally there were few.) While I was there, I also converted the
3040 str(n)cmp tests so they don't re-test the leading "-" and the first
3041 character, in the hope this might squeeze out yet more improvement.
3043 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3044 flag allowing group syntax was set by the header_syntax check but not
3045 turned off, possible causing trouble later; (2) The flag was not being
3046 set at all for the header_verify test, causing "group"-style headers to
3047 be rejected. I have now set it in this case, and also caused header_
3048 verify to ignore an empty address taken from a group. While doing this, I
3049 came across some other cases where the code for allowing group syntax
3050 while scanning a header line wasn't quite right (mostly, not resetting
3051 the flag correctly in the right place). These bugs could have caused
3052 trouble for malformed header lines. I hope it is now all correct.
3054 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3055 with the "reply" argument non-NULL. The code, however (which originally
3056 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3057 but it didn't always do it. This confused somebody who was copying the
3058 code for some other use. I have removed all the tests.
3060 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3061 feature that was used to support insecure browsers during the U.S. crypto
3062 embargo. It requires special client support, and Exim is probably the
3063 only MTA that supported it -- and would never use it because real RSA is
3064 always available. This code has been removed, because it had the bad
3065 effect of slowing Exim down by computing (never used) parameters for the
3066 RSA_EXPORT functionality.
3068 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3069 authenticator to fail if there's a tab character in the incoming data
3070 (there should never be unless someone is messing about, as it's supposed
3071 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3072 if the connection is using TLS or if the remote IP is the same as the
3073 local IP, and the "valid-client-cert option" if a client certificate has
3076 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3077 authenticators. This can be used for authorization after authentication
3078 succeeds. (In the case of plaintext, it servers for both authentication
3081 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3082 if any retry times were supplied.
3084 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3085 connection, where there is no client IP address to check. In this
3086 situation, the verify now always succeeds.
3088 PH/25 Applied John Jetmore's -Mset patch.
3090 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3092 PH/27 In a string expansion for a processed (not raw) header when multiple
3093 headers of the same name were present, leading whitespace was being
3094 removed from all of them, but trailing whitespace was being removed only
3095 from the last one. Now trailing whitespace is removed from each header
3096 before concatenation. Completely empty headers in a concatenation (as
3097 before) are ignored.
3099 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3100 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3102 PH/29 [Removed. This was a change that I later backed out, and forgot to
3103 correct the ChangeLog entry (that I had efficiently created) before
3104 committing the later change.]
3106 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3107 address errors (4xx response to RCPT) over the same connection as other
3108 messages routed to the same hosts. Such deliveries are always "forced",
3109 so retry times are not inspected. This resulted in far too many retries
3110 for the affected addresses. The effect occurred only when there were more
3111 hosts than the hosts_max_try setting in the smtp transport when it had
3112 the 4xx errors. Those hosts that it had tried were not added to the list
3113 of hosts for which the message was waiting, so if all were tried, there
3114 was no problem. Two fixes have been applied:
3116 (i) If there are any address or message errors in an SMTP delivery, none
3117 of the hosts (tried or untried) are now added to the list of hosts
3118 for which the message is waiting, so the message should not be a
3119 candidate for sending over the same connection that was used for a
3120 successful delivery of some other message. This seems entirely
3121 reasonable: after all the message is NOT "waiting for some host".
3122 This is so "obvious" that I'm not sure why it wasn't done
3123 previously. Hope I haven't missed anything, but it can't do any
3124 harm, as the worst effect is to miss an optimization.
3126 (ii) If, despite (i), such a delivery is accidentally attempted, the
3127 routing retry time is respected, so at least it doesn't keep
3128 hammering the server.
3130 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3131 in ${readsocket because some servers need this prod.
3133 PH/32 Added some extra debug output when updating a wait-xxx database.
3135 PH/33 The hint "could be header name not terminated by colon", which has been
3136 given for certain expansion errors for a long time, was not being given
3137 for the ${if def:h_colon_omitted{... case.
3139 PH/34 The spec says: "With one important exception, whenever a domain list is
3140 being scanned, $domain contains the subject domain." There was at least
3141 one case where this was not true.
3143 PH/35 The error "getsockname() failed: connection reset by peer" was being
3144 written to the panic log as well as the main log, but it isn't really
3145 panic-worthy as it just means the connection died rather early on. I have
3146 removed the panic log writing for the ECONNRESET error when getsockname()
3149 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3150 runs only) independently of the message's sender address. This meant
3151 that, if the 4xx error was in fact related to the sender, a different
3152 message to the same recipient with a different sender could confuse
3153 things. In particular, this can happen when sending to a greylisting
3154 server, but other circumstances could also provoke similar problems.
3155 I have changed the default so that the retry time for these errors is now
3156 based a combination of the sender and recipient addresses. This change
3157 can be overridden by setting address_retry_include_sender=false in the
3160 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3161 remote server are returned as part of bounce messages. This was not
3162 happening for LMTP over a pipe (the lmtp transport), but now it is the
3163 same for both kinds of LMTP.
3165 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3166 in header lines that were in fact CNAMEs. This is no longer the case.
3168 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3169 and queue runs started by the daemon processed all messages. This has
3170 been fixed so that -R and -S can now usefully be given with -q<time>.
3172 PH/40 Import PCRE release 6.7 (fixes some bugs).
3174 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3176 PH/42 Give an error if -q is specified more than once.
3178 PH/43 Renamed the variables $interface_address and $interface_port as
3179 $received_ip_address and $received_port, to make it clear that these
3180 values apply to message reception, and not to the outgoing interface when
3181 a message is delivered. (The old names remain recognized, of course.)
3183 PH/44 There was no timeout on the connect() call when using a Unix domain
3184 socket in the ${readsocket expansion. There now is.
3186 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3187 be meaningful with "accept".
3189 SC/01 Eximstats V1.43
3190 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3192 SC/02 Eximstats V1.44
3193 Use a glob alias rather than an array ref in the generated
3194 parser. This improves both readability and performance.
3196 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3197 Collect SpamAssassin and rejection statistics.
3198 Don't display local sender or destination tables unless
3199 there is data to show.
3200 Added average volumes into the top table text output.
3202 SC/04 Eximstats V1.46
3203 Collect data on the number of addresses (recipients)
3204 as well as the number of messages.
3206 SC/05 Eximstats V1.47
3207 Added 'Message too big' to the list of mail rejection
3208 reasons (thanks to Marco Gaiarin).
3210 SC/06 Eximstats V1.48
3211 Mainlog lines which have GMT offsets and are too short to
3212 have a flag are now skipped.
3214 SC/07 Eximstats V1.49 (Alain Williams)
3215 Added the -emptyok flag.
3217 SC/08 Eximstats V1.50
3218 Fixes for obtaining the IP address from reject messages.
3220 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3221 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3222 whitespace changes from 4.64-PH/27
3224 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3227 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3228 are found, allow negative numbers in numeric criteria)
3230 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3232 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3233 to match changes made in 4.64-PH/43
3235 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3237 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3238 "domain" is actually an IP address, and force a failure. This locks out
3239 those revolvers/nameservers that support "A-for-A" lookups, in
3240 contravention of the specifications.
3242 PH/48 When a host name was looked up from an IP address, and the subsequent
3243 forward lookup of the name timed out, the host name was left in
3244 $sender_host_name, contrary to the specification.
3246 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3247 restricted to single-key lookups, Exim was not diagnosing an error if
3248 * or *@ was used with a query-style lookup.
3250 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3252 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3253 long ago, but noone interested enough thought of it. Let's just say that
3254 the "1.1" means that there are some new functions that weren't there at
3255 some point in the past.
3257 PH/51 Error processing for expansion failure of helo_data from an smtp
3258 transport during callout processing was broken.
3260 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3261 tested/used via the -bh/-bhc/-bs options.
3263 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3264 bug, fixed in subsequent PCRE releases).
3266 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3267 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3269 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3275 SC/01 Use a glob alias rather than an array ref in eximstats generated
3276 parser. This improves both readability and performance.
3278 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3279 Don't display local sender or destination tables in eximstats unless
3280 there is data to show.
3281 Added average volumes into the eximstats top table text output.
3283 SC/03 Collect data on the number of addresses (recipients) as well
3284 as the number of messages in eximstats.
3286 TF/01 Correct an error in the documentation for the redirect router. Exim
3287 does (usually) call initgroups() when daemonizing.
3289 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3290 with consistent privilege compared to when running as a daemon.
3292 TF/03 Note in the spec that $authenticated_id is not set for local
3293 submissions from trusted users.
3295 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3296 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3298 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3299 by adding some example configuration directives to the default
3300 configuration file. A little bit of work is required to uncomment the
3301 directives and define how usernames and passwords are checked, but
3302 there is now a framework to start from.
3304 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3305 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3306 without this. I don't know how relevant this is to other LDAP libraries.
3308 PH/02 Add the verb name to the "unknown ACL verb" error.
3310 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3312 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3314 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3315 directory not expanded when it should be if an expanded home directory
3316 was set for the address (which is overridden by the transport).
3318 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3321 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3322 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3323 because it is too late at that time, and has no effect.
3325 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3326 security issue with \' (bugzilla #107). I could not use the
3327 PQescapeStringConn() function, because it needs a PGconn value as one of
3330 PH/08 When testing addresses using -bt, indicate those final addresses that
3331 are duplicates that would not cause an additional delivery. At least one
3332 person was confused, thinking that -bt output corresponded to deliveries.
3333 (Suppressing duplicates isn't a good idea as you lose the information
3334 about possibly different redirections that led to the duplicates.)
3336 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3337 systems where poll() doesn't work, in particular OS X.
3339 PH/10 Added more information to debugging output for retry time not reached.
3341 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3342 operations in malware.c.
3344 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3347 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3348 syslog_duplication set false, log lines that would normally be written
3349 both the the main log and to the reject log were not written to syslog at
3352 PH/14 In the default configuration, change the use of "message" in ACL warn
3353 statements to "add_header".
3355 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3356 not followed by a command (e.g. "seen endif").
3358 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3359 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3362 PH/17 Added extra conditions to the default value of delay_warning_condition
3366 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3367 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3368 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3371 The Auto-Submitted: and various List- headers are standardised, whereas I
3372 don't think Precedence: ever was.
3374 PH/18 Refactored debugging code in route_finduser() to show more information,
3375 in particular, the error code if getpwnam() issues one.
3377 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3378 This is apparently needed in addition to the PH/07 change above to avoid
3379 any possible encoding problems.
3381 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3382 but not after initializing Perl.
3384 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3385 output them only if debugging. By default they are written stderr,
3386 apparently, which is not desirable.
3388 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3391 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3394 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3396 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3397 authenticated or an ident call has been made. Suppress the default
3398 values for $authenticated_id and $authenticated_sender (but permit -oMai
3399 and -oMas) when testing with -bh.
3401 PH/24 Re-jigged the order of the tests in the default configuration so that the
3402 tests for valid domains and recipients precede the DNS black list and CSA
3403 tests, on the grounds that those ones are more expensive.
3405 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3406 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3407 This bug exists in every version of Exim that I still have, right back to
3410 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3411 However, an attempt to turn on case-sensitivity in a regex key by
3412 including (?-i) didn't work because the subject string was already
3413 lowercased, and the effects were non-intuitive. It turns out that a
3414 one-line patch can be used to allow (?-i) to work as expected.
3420 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3421 other effects) broke the use of negated acl sub-conditions.
3423 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3426 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3427 "Deny" causes Exim to reject the incoming connection with a 554 error.
3428 Unfortunately, if there is a major crisis, such as a disk failure,
3429 tcp-wrappers gives "deny", whereas what one would like would be some
3430 kind of temporary error. A kludge has been added to help with this.
3431 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3432 554 error is used if errno is still zero or contains ENOENT (which occurs
3433 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3436 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3438 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3439 errors. Otherwise a message that provokes a temporary error (when other
3440 messages do not) can cause a whole host to time out.
3442 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3443 addresses were routed directly to files or pipes from a redirect router.
3444 File deliveries just didn't batch; pipe deliveries might have suffered
3447 PH/06 A failure to get a lock for a hints database would erroneously always say
3448 "Failed to get write lock", even when it was really a read lock.
3450 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3451 of 0600. This has been changed to use the value of the lockfile_mode
3452 option (which defaults to 0600).
3454 PH/08 Applied small patch from the Sieve maintainer.
3456 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3457 folder from quota calculations, a direct delivery into this folder messed
3458 up the contents of the maildirsize file. This was because the regex was
3459 used only to exclude .Trash (or whatever) when the size of the mailbox
3460 was calculated. There was no check that a delivery was happening into an
3461 excluded directory. This bug has been fixed by ignoring all quota
3462 processing for deliveries into excluded directories.
3464 PH/10 Added the maildirfolder_create_regex option to appendfile.
3470 PH/01 The code for finding all the local interface addresses on a FreeBSD
3471 system running IPv6 was broken. This may well have applied to all BSD
3472 systems, as well as to others that have similar system calls. The broken
3473 code found IPv4 interfaces correctly, but gave incorrect values for the
3474 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3475 that it would not match correctly against @[] and not recognize the IPv6
3478 PH/02 The ipliteral router was not recognizing addresses of the form user@
3479 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3481 PH/03 Added disable_ipv6.
3483 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3484 decoded form, because it is most often used to construct To: headers
3485 lines in autoreplies, and the decoded form may well be syntactically
3486 invalid. However, $reply_address has leading white space removed, and all
3487 newlines turned into spaces so that the autoreply transport does not
3490 PH/05 If group was specified without a user on a router, and no group or user
3491 was specified on a transport, the group from the router was ignored.
3493 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3494 for visible compile-time settings that can be used to change these
3495 numbers, for those that want even more. Backwards compatibility with old
3496 spool files has been maintained. However, going back to a previous Exim
3497 release will lost any variables that are in spool files.
3499 PH/07 Two small changes when running in the test harness: increase delay when
3500 passing a TCP/IP connection to a new process, in case the original
3501 process has to generate a bounce, and remove special handling of
3502 127.0.0.2 (sic), which is no longer necessary.
3504 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3505 be the same on different OS.
3507 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3510 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3511 whether --show-vars was specified or not
3513 JJ/02 exipick: Added support for new ACL variable spool format introduced
3516 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3517 syntactically invalid From: or Reply-to: line, and a filter used this to
3518 generate an autoreply, and therefore failed to obtain an address for the
3519 autoreply, Exim could try to deliver to a non-existent relative file
3520 name, causing unrelated and misleading errors. What now happens is that
3521 it logs this as a hard delivery error, but does not attempt to create a
3524 PH/11 The exinext utility has a -C option for testing purposes, but although
3525 the given file was scanned by exinext itself; it wasn't being passed on
3526 when Exim was called.
3528 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3529 an end-of-file indication when reading a command response.
3531 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3532 compiled. In many other places in Exim, IPv6 addresses are always
3533 recognized, so I have changed this. It also means that IPv4 domain
3534 literals of the form [IPV4:n.n.n.n] are now always recognized.
3536 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3537 used if the router is not running as root, for example, when verifying at
3538 ACL time, or when using -bh. The debugging output from this situation was
3539 non-existent - all you got was a failure to exec. I have made two
3542 (a) Failures to set uid/gid, the current directory, or a process leader
3543 in a subprocess such as that created by queryprogram now generate
3544 suitable debugging output when -d is set.
3546 (b) The queryprogram router detects when it is not running as root,
3547 outputs suitable debugging information if -d is set, and then runs
3548 the subprocess without attempting to change uid/gid.
3550 PH/15 Minor change to Makefile for building test_host (undocumented testing
3553 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3554 additional section of a DNS packet that returns MX or SRV records.
3555 Instead, it always explicitly searches for A/AAAA records. This avoids
3556 major problems that occur when a DNS server includes only records of one
3557 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3558 fixed another bug: if SRV records were looked up and the corresponding
3559 address records were *not* found in the additional section, the port
3560 values from the SRV records were lost.
3562 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3563 using the correct key (the original address) when searching the retry
3564 rules in order to find which one to use for generating the retry hint.
3566 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3567 adding the default one. Similarly, if it contains a Reply-To: header, the
3568 errors_reply_to option, if set, is not used.
3570 PH/19 When calculating a retry time, Exim used to measure the "time since
3571 failure" by looking at the "first failed" field in the retry record. Now
3572 it does not use this if it is later than than the arrival time of the
3573 message. Instead it uses the arrival time. This makes for better
3574 behaviour in cases where some deliveries succeed, thus re-setting the
3575 "first failed" field. An example is a quota failure for a huge message
3576 when small messages continue to be delivered. Without this change, the
3577 "time since failure" will always be short, possible causing more frequent
3578 delivery attempts for the huge message than are intended.
3579 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3581 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3582 $1, $2, $3) because the numerical variables can be reset during some
3583 expansion items (e.g. "match"), thereby losing the authentication data.
3585 PH/21 Make -bV show the size of off_t variables so that the test suite can
3586 decide whether to run tests for quotas > 2G.
3588 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3589 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3590 filecount value is greater than 2G or if a quota value is greater than 2G
3591 on a system where the size of off_t is not greater than 4, a panic error
3594 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3595 never match. The debug and -bh output now contains an explicit error
3596 message indicating a malformed IPv4 address or mask.
3598 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3599 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3600 PH/23 above applies.
3602 PH/25 Do not write to syslog when running in the test harness. The only
3603 occasion when this arises is a failure to open the main or panic logs
3604 (for which there is an explicit test).
3606 PH/26 Added the /no_tell option to "control=freeze".
3608 PH/27 If a host name lookup failed very early in a connection, for example, if
3609 the IP address matched host_lookup and the reverse lookup yielded a name
3610 that did not have a forward lookup, an error message of the form "no IP
3611 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3612 could be logged. Now it outputs the IP address instead of "NULL".
3614 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3615 allows the sender and the authenticated sender to be set when
3616 submitting a message from within Exim. Since child_open_exim() is
3617 documented for local_scan(), the new function should be too.
3619 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3620 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3621 results in an empty string is now treated as unset.
3623 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3625 PH/31 Added sender_verify_fail logging option.
3627 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3628 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3629 tidied the source and removed it altogether.
3631 PH/33 When a queue run was abandoned because the load average was too high, a
3632 log line was always written; now it is written only if the queue_run log
3633 selector is set. In addition, the log line for abandonment now contains
3634 information about the queue run such as the pid. This is always present
3635 in "start" and "stop" lines but was omitted from the "abandon" line.
3637 PH/34 Omit spaces between a header name and the colon in the error message that
3638 is given when verify = headers_syntax fails (if there are lots of them,
3639 the message gets confusing).
3641 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3642 names, as there are now some PTR records that contain slashes. This check
3643 is only to protect against broken name servers that fall over on strange
3644 characters, so the fact that it applies to all lookups doesn't matter.
3646 PH/36 Now that the new test suite is complete, we can remove some of the
3647 special code in Exim that was needed for the old test suite. For example,
3648 sorting DNS records because real resolvers return them in an arbitrary
3649 order. The new test suite's fake resolver always returns records in the
3652 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3653 bounces) except when queue_only is set, to avoid logging races between
3654 the different processes.
3656 PH/38 Panic-die if .include specifies a non-absolute path.
3658 PH/39 A tweak to the "H" retry rule from its user.
3660 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3661 a label. They prevented compilation on older perls.
3663 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3664 a warning to be raised on newish perls.
3666 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3667 on queue. Changes to match documented behaviour of showing count of
3668 messages matching specified criteria.
3670 PH/40 Changed the default ident timeout from 30s to 5s.
3672 PH/41 Added support for the use of login_cap features, on those BSD systems
3673 that have them, for controlling the resources used by pipe deliveries.
3675 PH/42 The content-scanning code uses fopen() to create files in which to put
3676 message data. Previously it was not paying any attention to the mode of
3677 the files. Exim runs with umask(0) because the rest of the code creates
3678 files with open(), and sets the required mode explicitly. Thus, these
3679 files were ending up world-writeable. This was not a big issue, because,
3680 being within the spool directory, they were not world-accessible. I have
3681 created a function called modefopen, which takes an additional mode
3682 argument. It sets umask(777), creates the file, chmods it to the required
3683 mode, then resets the umask. All the relevant calls to fopen() in the
3684 content scanning code have been changed to use this function.
3686 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3687 to 24 hours. This avoids potential overflow problems when processing G
3688 and H retry rules. I suspect nobody ever tinkers with this value.
3690 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3692 PH/45 When the plaintext authenticator is running as a client, the server's
3693 challenges are checked to ensure they are valid base64 strings. By
3694 default, the authentication attempt is cancelled if an invalid string is
3695 received. Setting client_ignore_invalid_base64 true ignores these errors.
3696 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3697 they are received. Thus, the responses can be made to depend on the
3698 challenges. If an invalid string is ignored, an empty string is placed in
3701 PH/46 Messages that are created by the autoreply transport now contains a
3702 References: header, in accordance with RFCs 2822 and 3834.
3704 PH/47 Added authenticated_sender_force to the smtp transport.
3706 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3708 PH/49 Installed latest patch from the Sieve maintainer.
3710 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3711 was also set, the appendfile transport was unnecessarily scanning a
3712 directory of message files (e.g. for maildir delivery) to find the count
3713 of files (along with the size), even though it did not need this
3714 information. It now does the scan only if it needs to find either the
3715 size of the count of files.
3717 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3719 PH/52 Two bugs concerned with error handling when the smtp transport is
3722 (i) Exim was not creating retry information for temporary errors given
3723 for individual recipients after the DATA command when the smtp transport
3724 was used in LMTP mode. This meant that they could be retried too
3725 frequently, and not timed out correctly.
3727 (ii) Exim was setting the flag that allows error details to be returned
3728 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3729 recipients that were returned after the DATA command.
3731 PH/53 This is related to PH/52, but is more general: for any failing address,
3732 when detailed error information was permitted to be returned to the
3733 sender, but the error was temporary, then after the final timeout, only
3734 "retry timeout exceeded" was returned. Now it returns the full error as
3735 well as "retry timeout exceeded".
3737 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3738 do this, and (what is worse) MTAs that accept it.
3740 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3741 will now be deprecated.
3743 PH/56 New os.c-cygwin from the Cygwin maintainer.
3745 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3746 formats (previously only available in exim formats via -bpr, -bpru,
3747 and -bpra. Now also available in native and exiqgrep formats)
3749 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3750 with very large, slow to parse queues
3752 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3754 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3756 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3757 responses to authentication challenges, though it was showing the
3758 challenges; (ii) I've removed the CR characters from the debug output for
3761 PH/58 Allow for the insertion of a newline as well as a space when a string
3762 is turned into more than one encoded-word during RFC 2047 encoding. The
3763 Sieve code now uses this.
3765 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3766 data_4xx, lost_connection, tls_required.
3768 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3769 message was being sent as an SMTP response.
3771 PH/61 Add -l and -k options to exicyclog.
3773 PH/62 When verifying, if an address was redirected to one new address, so that
3774 verification continued, and the new address failed or deferred after
3775 having set something in $address_data, the value of $address_data was not
3776 passed back to the ACL. This was different to the case when no
3777 redirection occurred. The value is now passed back in both cases.
3779 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3780 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3781 use login_cap.h, so on its own it isn't the distinguishing feature. The
3782 new name refers directly to the setclassresources() function.
3784 PH/65 Added configuration files for NetBSD3.
3786 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3788 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3789 is preferred over IPv4.
3791 PH/68 The bounce_return_message and bounce_return_body options were not being
3792 honoured for bounces generated during the reception of non-SMTP messages.
3793 In particular, this applied to messages rejected by the ACL. This bug has
3794 been fixed. However, if bounce_return_message is true and bounce_return_
3795 body is false, the headers that are returned for a non-SMTP message
3796 include only those that have been read before the error was detected.
3797 (In the case of an ACL rejection, they have all been read.)
3799 PH/69 The HTML version of the specification is now built in a directory called
3800 spec_html instead of spec.html, because the latter looks like a path with
3801 a MIME-type, and this confuses some software.
3803 PH/70 Catch two compiler warnings in sieve.c.
3805 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3806 function verify_get_ident() calls ip_connect() to connect a socket, but
3807 if the "connect()" function timed out, ip_connect() used to close the
3808 socket. However, verify_get_ident() also closes the socket later, and in
3809 between Exim writes to the log, which may get opened at this point. When
3810 the socket was closed in ip_connect(), the log could get the same file
3811 descriptor number as the socket. This naturally causes chaos. The fix is
3812 not to close the socket in ip_connect(); the socket should be closed by
3813 the function that creates it. There was only one place in the code where
3814 this was missing, in the iplookup router, which I don't think anybody now
3815 uses, but I've fixed it anyway.
3817 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3818 well as to direct DNS lookups. Otherwise the handling of names in host
3819 lists is inconsistent and therefore confusing.
3825 PH/01 Two changes to the default runtime configuration:
3827 (1) Move the checks for relay_from_hosts and authenticated clients from
3828 after to before the (commented out) DNS black list checks.
3830 (2) Add control=submission to the relay_from_hosts and authenticated
3831 clients checks, on the grounds that messages accepted by these
3832 statements are most likely to be submissions.
3834 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3836 (1) Generate an error if the third argument for the ${prvs expansion is
3839 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3842 (3) Reset the variables that are obtained from the first argument of
3843 ${prvscheck and used in the second argument before leaving the code,
3844 because their memory is reclaimed, so using them afterwards may do
3847 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3848 one (it's much easier than Tom thought :-).
3850 (5) Because of (4), we can now allow for the use of $prvscheck_result
3851 inside the third argument.
3853 PH/03 For some reason, the default setting of PATH when running a command from
3854 a pipe transport was just "/usr/bin". I have changed it to
3857 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3858 anything to be listed in the output from -bV.
3860 PH/05 When a filter generated an autoreply, the entire To: header line was
3861 quoted in the delivery log line, like this:
3863 => >A.N.Other <ano@some.domain> <original@ddress> ...
3865 This has been changed so that it extracts the operative address. There
3866 may be more than one such address. If so, they are comma-separated, like
3869 => >ano@some.domain,ona@other.domain <original@ddress> ...
3871 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3872 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3873 not being looked up in the rDNS to get a host name, Exim was showing the
3874 IP address twice in Received: lines, even though the IP addresses were
3875 identical. For example:
3877 Received: from [1.2.3.4] (helo=[1.2.3.4])
3879 However, if the real host name was known, it was omitting the HELO data
3880 if it matched the actual IP address. This has been tidied up so that it
3881 doesn't show the same IP address twice.
3883 PH/07 When both +timestamp and +memory debugging was on, the value given by
3884 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3885 called by the debug printing, thereby overwriting the timestamp buffer.
3886 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3888 PH/08 When the original message was included in an autoreply transport, it
3889 always said "this is a copy of the message, including all the headers",
3890 even if body_only or headers_only was set. It now gives an appropriate
3893 PH/09 Applied a patch from the Sieve maintainer which:
3895 o fixes some comments
3896 o adds the (disabled) notify extension core
3897 o adds some debug output for the result of if/elsif tests
3898 o points to the current vacation draft in the documentation
3899 and documents the missing references header update
3903 o fixes a bug in processing the envelope test (when testing
3904 multiple envelope elements, the last element determined the
3907 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3908 Electronic Mail") by including:
3910 Auto-submitted: auto-generated
3912 in the messages that it generates (bounce messages and others, such as
3913 warnings). In the case of bounce messages for non-SMTP messages, there was
3914 also a typo: it was using "Auto_submitted" (underscore instead of
3915 hyphen). Since every message generated by Exim is necessarily in response
3916 to another message, thes have all been changed to:
3918 Auto-Submitted: auto-replied
3920 in accordance with these statements in the RFC:
3922 The auto-replied keyword:
3924 - SHOULD be used on messages sent in direct response to another
3925 message by an automatic process,
3927 - MUST NOT be used on manually-generated messages,
3929 - MAY be used on Delivery Status Notifications (DSNs) and Message
3930 Disposition Notifications (MDNs),
3932 - MUST NOT be used on messages generated by automatic or periodic
3933 processes, except for messages which are automatic responses to
3936 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3937 to the default Received: header definition.
3939 PH/12 Added log selector acl_warn_skipped (default on).
3941 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3942 variables because (a) they are in the wrong storage pool and (b) even if
3943 they were copied, it wouldn't work properly because of the caching.
3945 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3946 checking when decoding. Apparently there are clients that generate
3947 overlong encoded strings. Why am I not surprised?
3949 PH/15 If the first argument of "${if match_address" was not empty, but did not
3950 contain an "@" character, Exim crashed. Now it writes a panic log message
3951 and treats the condition as false.
3953 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3955 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3956 extension "envelope-auth". The code is finished and in agreement with
3957 other implementations, but there is no documentation so far and in fact,
3958 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3959 not changing the active code.
3961 Print executed "if" and "elsif" statements when debugging is used. This
3962 helps a great deal to understand what a filter does.
3964 Document more things not specified clearly in RFC3028. I had all this
3965 sorted out, when out of a sudden new issues came to my mind. Oops."
3967 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3970 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3971 canonical form (as documented). However, after a host name lookup from
3972 the IP address, check_host() was doing a simple string comparison with
3973 addresses acquired from the DNS when checking that the found name did
3974 have the original IP as one of its addresses. Since any found IPv6
3975 addresses are likely to be in abbreviated form, the comparison could
3976 fail. Luckily, there already exists a function for doing the comparison
3977 by converting both addresses to binary, so now that is used instead of
3978 the text comparison.
3980 PH/20 There was another similar case to PH/19, when a complete host name was
3981 given in a host list; looking up its IP address could give an abbreviated
3982 form, whereas the current host's name might or might not be abbreviated.
3983 The same fix has been applied.
3989 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3990 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3993 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3994 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3996 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3998 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3999 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4000 Also, for "Auto-Submitted", treat anything other than "no" as
4001 non-personal, in accordance with RFC 3834. (Previously it treated
4002 anything starting "auto-" as non-personal.)
4004 TF/01 The control=submission/name=... option had a problem with syntax
4005 errors if the name included a slash character. The /name= option
4006 now slurps the rest of the string, so it can include any characters
4007 but it must come last in the list of options (after /sender_retain
4010 PH/05 Some modifications to the interface to the fake nameserver for the new
4018 TK/01 Added the "success_on_redirect" address verification option. See
4019 NewStuff for rationale and an example.
4021 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4023 PH/02 Patch to exigrep to allow it to work on syslog lines.
4025 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4026 fread() to skip over the body file's header line, because in Cygwin the
4027 header line is locked and is inaccessible.
4029 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4030 co-exist for some time) to make it clear that it is the Exim ID that is
4031 referenced, not the Message-ID: header line.
4033 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4034 string_format() function, because snprintf() does not exist on all
4037 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4038 ${stat: expansion item.
4040 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4041 protocol synchronization error", to keep the pedants happy.
4043 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4044 well as for IRIX systems, when gcc is being used. See the host.c source
4047 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4049 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4052 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4053 transport and to the smtp transport in LMTP mode.
4055 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4057 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4058 run a filter in a subprocess. This could lead to confusion in subsequent
4059 lookups in the parent process. There should also be a search_tidyup() at
4060 the end of the subprocess.
4062 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4063 only if the host matched helo_try_verify_hosts, which caused the
4064 verification to occur when the EHLO/HELO command was issued. The ACL just
4065 tested the remembered result. Now, if a previous verification attempt has
4066 not happened, "verify = helo" does it there and then.
4068 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4070 TK/03 Fix log output including CR from clamd.
4072 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4073 exist provoked a memory error which could cause a segfault.
4075 PH/15 Installed PCRE 6.2
4077 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4079 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4080 of the problem. Specifically, suggested +O2 rather than +O1 for the
4083 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4085 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4086 record, the port was not correctly logged when the outgoing_port log
4087 selector was set (it logged the transort's default port).
4089 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4090 fallback_hosts, and "hosts" in the smtp transport.
4092 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4093 host errors such as "Connection refused".
4095 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4096 authentication with radiusclient 0.4.9:
4098 - Error returned from rc_read_config was caught wrongly
4099 - Username/password not passed on to radius server due to wrong length.
4101 The presumption is that some radiusclient API changes for 4.51/PH/17
4102 were not taken care of correctly. The code is still untested by me (my
4103 Linux distribution still has 0.3.2 of radiusclient), but it was
4104 contributed by a Radius user.
4106 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4107 expanding the "port" option of the smtp transport.
4109 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4110 while reading a MIME header. Thanks to Tom Hughes for a patch.
4112 PH/24 Include config.h inside local_scan.h so that configuration settings are
4115 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4116 that in an ACL for RCPT (for example), you can examine exactly what was
4119 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4120 commands, but it was not correctly comparing the address with the actual
4121 client host address. Thus, it would show the EHLO address in Received:
4122 header lines when this was not necessary.
4124 PH/27 Added the % operator to ${eval:}.
4126 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4127 it should be ignoring failures (because with -C, for example, it has lost
4128 privilege). It wasn't ignoring creation failures other than "already
4131 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4134 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4135 because an input line was too long, either on its own, or by virtue of
4136 too many continuations, the temporary file was not being removed, and the
4137 return code was incorrect.
4139 PH/31 Missing "BOOL" in function definition in filtertest.c.
4141 PH/32 Applied Sieve patches from the maintainer.
4143 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4145 PH/33 Added "verify = not_blind".
4147 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4148 Local/Makefile (with some defaults set). These are used in built scripts
4149 such as exicyclog, but they have never been used in the exim_install
4150 script (though there are many overriding facilities there). I have
4151 arranged that the exim_install script now takes note of these two
4154 PH/35 Installed configuration files for Dragonfly.
4156 PH/36 When a locally submitted message by a trusted user did not contain a
4157 From: header, and the sender address was obtained from -f or from an SMTP
4158 MAIL command, and the trusted user did not use -F to supply a sender
4159 name, $originator_name was incorrectly used when constructing a From:
4160 header. Furthermore, $originator_name was used for submission mode
4161 messages from external hosts without From: headers in a similar way,
4162 which is clearly wrong.
4164 PH/37 Added control=suppress_local_fixups.
4166 PH/38 When log_selector = +received_sender was set, and the addition of the
4167 sender made the log line's construction buffer exactly full, or one byte
4168 less than full, an overflow happened when the terminating "\n" was
4171 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4172 when the result of a list match is failure because a DNS lookup failed.
4174 PH/40 RM_COMMAND is now used in the building process.
4176 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4177 the "build-* directories that it finds.
4179 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4180 address in a domain literal was a prefix of an interface address.
4182 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4183 when verifying a sender address, unless rewrite_headers is false.
4185 PH/44 Wrote a long comment about why errors_to addresses are verified as
4186 recipients, not senders.
4188 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4189 the ratelimit ACL was added.
4191 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4193 PH/46 Added extra information about PostgreSQL errors to the error string.
4195 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4196 avoiding the need to install special zones in a real server. This is
4197 backwards compatible; if it can't find the fake resolver, it drops back.
4198 Thus, both old and new test suites can be run.
4200 TF/02 Added util/ratelimit.pl
4202 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4203 clock is set back in time.
4205 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4206 Candler <B.Candler@pobox.com>.
4208 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4209 OK for addresses that are the result of redirections.
4211 PH/48 A number of further additions for the benefit of the new test suite,
4212 including a fake gethostbyname() that interfaces to the fake DNS resolver
4215 TF/06 The fix for widen_domains has also been applied to qualify_single and
4216 search_parents which are the other dnslookup options that can cause
4219 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4222 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4224 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4225 service" if it encountered an MX record with an empty target hostname.
4226 The message is now "an MX or SRV record indicated no SMTP service".
4228 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4229 if the DNS of the sending site is misconfigured. This is quite a
4230 common situation. This change restores the behaviour of treating a
4231 helo verification defer as a failure.
4233 PH/51 If self=fail was set on a router, the bounce message did not include the
4234 actual error message.
4240 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4242 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4243 command itself ended in error, the underlying message about the transport
4244 filter timeout was being overwritten with the pipe command error. Now the
4245 underlying error message should be appended to the second error message.
4247 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4249 PH/02 Reduce the amount of output that "make" produces by default. Full output
4250 can still be requested.
4252 PH/03 The warning log line about a condition test deferring for a "warn" verb
4253 was being output only once per connection, rather than after each
4254 occurrence (because it was using the same function as for successful
4255 "warn" verbs). This seems wrong, so I have changed it.
4257 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4258 it should not have, which might have caused a crash in the right
4259 circumstances, but probably never did.
4261 PH/04 Installed a modified version of Tony Finch's patch to make submission
4262 mode fix the return path as well as the Sender: header line, and to
4263 add a /name= option so that you can make the user's friendly name appear
4266 TF/03 Added the control = fakedefer ACL modifier.
4268 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4269 Mark Lowes for thorough testing.
4271 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4273 TK/03 Merged latest SRS patch from Miles Wilton.
4275 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4276 with the definition in sysexits.h (which is #included earlier).
4277 Fortunately, Exim does not actually use EX_OK. The code used to try to
4278 preserve the sysexits.h value, by assuming that macro definitions were
4279 scanned for macro replacements. I have been disabused of this notion,
4280 so now the code just undefines EX_OK before #including unistd.h.
4282 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4283 in the smtp transport. When a block could not be written in a single
4284 write() function, the timeout was being re-applied to each part-write.
4285 This seems wrong - if the receiver was accepting one byte at a time it
4286 would take for ever. The timeout is now adjusted when this happens. It
4287 doesn't have to be particularly precise.
4289 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4290 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4292 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4293 without a domain if the check to <postmaster@domain> fails.
4295 SC/01 Eximstats: added -xls and the ability to specify output files
4296 (patch written by Frank Heydlauf).
4298 SC/02 Eximstats: use FileHandles for outputting results.
4300 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4302 SC/04 Eximstats: fixed display of large numbers with -nvr option
4304 SC/05 Eximstats: fixed merging of reports with empty tables.
4306 SC/06 Eximstats: added the -include_original_destination flag
4308 SC/07 Eximstats: removed tabs and trailing whitespace.
4310 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4312 TK/06 MBOX spool code: Add real "From " MBOX separator line
4313 so the .eml file is really in mbox format (even though
4314 most programs do not really care). Patch from Alex Miller.
4316 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4317 The latter is generated from $received_to and is only set if the
4318 message has one envelope recipient. SA can use these headers,
4319 obviously out-of-the-box. Patch from Alex Miller.
4321 PH/08 The ${def test on a variable was returning false if the variable's
4322 value was "0", contrary to what the specification has always said!
4323 The result should be true unless the variable is empty.
4325 PH/09 The syntax error of a character other than { following "${if
4326 def:variable_name" (after optional whitespace) was not being diagnosed.
4327 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4328 accidental colon was present, for example, could give incorrect results.
4330 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4331 result is used (not including appendfile, where other changes are about
4334 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4335 This involved changing a lot of size variables from int to off_t. It
4336 should work with maildirs and everything.
4338 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4339 spamd dying while we are connected to it.
4341 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4344 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4345 function for BDB 4.3.
4347 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4349 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4350 the sources for PCRE have been re-arranged and more files are now
4353 PH/15 The code I had for printing potentially long long variables in PH/11
4354 above was not the best (it lost precision). The length of off_t variables
4355 is now inspected at build time, and an appropriate printing format (%ld
4356 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4357 to be "long long int" or "long int". This is needed for the internal
4358 formatting function string_vformat().
4360 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4361 the configuration file to be ":syslog", then the script "guesses" where
4362 the logs files are, rather than using the compiled in default. In our
4363 case the guess is not the same as the compiled default, so the script
4364 suddenly stopped working when I started to use syslog. The patch checks
4365 to see if log_file_path is "". If so, it attempts to read it from exim
4366 with no configuration file to get the compiled in version, before it
4367 falls back to the previous guessing code."
4369 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4370 implementing BATV in an Exim configuration. See NewStuff for the gory
4373 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4374 Makefile that are specific to HP-UX.
4376 PH/18 If the "use_postmaster" option was set for a recipient callout together
4377 with the "random" option, the postmaster address was used as the MAIL
4378 FROM address for the random test, but not for the subsequent recipient
4379 test. It is now used for both.
4381 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4382 patch removes a few documentation additions to RFC 3028, because the
4383 latest draft now contains them. It adds the new en;ascii-case comparator
4384 and a new error check for 8bit text in MIME parts. Comparator and
4385 require names are now matched exactly. I enabled the subaddress
4386 extension, but it is not well tested yet (read: it works for me)."
4388 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4389 rework some of the code of TK/09 above to avoid the hardwired use of
4390 "%lld" and "long long". Replaced the call to snprintf() with a call to
4393 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4394 records point to non-existent hosts", "retry timeout exceeded", and
4395 "retry time not reached for any host after a long failure period".
4397 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4398 experimental DomainKeys support:
4400 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4401 (2) On an error such as an illegally used "control", the wrong name for
4402 the control was given.
4404 These problems did NOT occur unless DomainKeys support was compiled.
4406 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4408 PH/24 Added ${if match_ip condition.
4410 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4411 databases so that it will be absolutely obvious if a crash occurs in the
4412 DB library. This is a regular occurrence (often caused by mis-matched
4415 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4416 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4417 picked up on a user's system that detects such things. There doesn't seem
4418 to be a gcc warning option for this - only an attribute that has to be
4419 put on the function's prototype. It seems that in Fedora Core 4 they have
4420 set this on a number of new functions. No doubt there will be more in due
4423 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4424 specify a transport. However, if an address that was verified by such a
4425 router was the subject of a callout, Exim crashed because it tried to
4426 read the rcpt_include_affixes from the non-existent transport. Now it
4427 just assumes that the setting of that option is false. This bug was
4428 introduced by 4.51/PH/31.
4430 PH/28 Changed -d+all to exclude +memory, because that information is very
4431 rarely of interest, but it makes the output a lot bigger. People tend to
4432 do -d+all out of habit.
4434 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4435 code in os-type was giving problems when libc.so lives in lib64, like on
4438 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4439 aren't the modern standard, and it seems that some systems' include files
4440 don't always have them. Exim was already checking for some of the newer
4441 ones like T_AAAA, and defining it itself. I've added checks for all the
4442 record types that Exim uses.
4444 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4445 not automatically generating a new one, as it is supposed to. This
4446 prevented TLS from working. If the file did exist, but contained invalid
4447 data, a new version was generated, as expected. It was only the case of a
4448 non-existent file that was broken.
4450 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4451 with a change in libdomainkeys > 0.64.
4453 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4454 from DNS. If the selector record carries the flag, it now has
4455 precedence over the domain-wide flag.
4457 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4459 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4460 the use of a port name that isn't defined in /etc/services) occurred, the
4461 message was deferred as in a normal delivery, and thus remained on the
4462 spool, instead of being failed because of the mua_wrapper setting. This
4463 is now fixed, and I tidied up some of the mua_wrapper messages at the
4466 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4467 the messages in a hash of arrays rather than using individual hashes.
4468 This is a bit cleaner and results in dramatic memory savings, albeit
4469 at a slight CPU cost.
4471 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4472 as requested by Marc Sherman.
4474 SC/10 Eximstats: added histograms for user specified patterns as requested
4477 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4479 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4480 fopen() in the content-scanning modules that did not already have it.
4486 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4487 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4489 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4491 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4493 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4494 to test Sieve filters that use "vacation".
4496 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4497 that changes the way the GnuTLS parameters are stored in the cache file.
4498 The new format can be generated externally. For backward compatibility,
4499 if the data in the cache doesn't make sense, Exim assumes it has read an
4500 old-format file, and it generates new data and writes a new file. This
4501 means that you can't go back to an older release without removing the
4504 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4505 work if there are any delivery delays because "one_time" forces the
4506 parent to be marked "delivered", so its unseen clone is never tried
4507 again. For this reason, Exim now forbids the simultaneous setting of
4510 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4511 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4512 entry if you want to know the details. The fix, however, neglected to
4513 consider the case when local delivery batching is involved. The test for
4514 "previously delivered" was not happening when checking to see if an
4515 address could be batched with a previous (undelivered) one; under
4516 certain circumstances this could lead to multiple deliveries to the same
4519 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4520 in its include files, and this causes problems building Exim.
4522 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4523 header_syntax) but Exim was just ignoring anything given after a slash.
4524 In particular, this caused confusion with an attempt to use "verify =
4525 reverse_host_lookup/defer_ok". An error is now given when options are
4526 supplied for verify items that do not have them. (Maybe reverse_host_
4527 lookup should have a defer_ok option, but that's a different point.)
4529 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4530 defined by RFC 821) to 2048, because there were problems with some AUTH
4531 commands, and RFC 1869 says the size should be increased for extended
4532 SMTP commands that take arguments.
4534 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4537 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4538 "unknown" error; now it says that the functionality isn't in the binary.
4540 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4541 an address' error message when a string expansion fails (syntax or
4542 whatever). Otherwise the password may appear in the log. Following change
4543 PH/42 below, there is no longer a chance of it appearing in a bounce
4546 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4548 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4549 of its addresses was ever tried. (Bugzilla bug #2.)
4551 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4552 the result incorrectly in the debug output. (It correctly added a newline
4553 to what was transported.)
4555 TF/01 Added $received_time.
4557 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4558 commented out examples of how to interface to a virus scanner and to
4559 SpamAssassin. Also added commented examples of av_scanner and
4560 spamd_address settings.
4562 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4563 and controls are allowed in which ACLs. There were a couple of minor
4564 errors. Some of the entries in the conditions table (which is a table of
4565 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4566 negation of where the condition IS allowed.
4568 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4570 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4571 header file does not have a version number, so I've had to invent a new
4572 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4573 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4574 radiusclient), but it was contributed by a Radius user.
4576 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4577 files or directories, for OpenSSL.
4579 PH/19 When an Exim process that is running as root has to create an Exim log
4580 file, it does so in a subprocess that runs as exim:exim so as to get the
4581 ownership right at creation (otherwise, other Exim processes might see
4582 the file with the wrong ownership). There was no test for failure of this
4583 fork() call, which would lead to the process getting stuck as it waited
4584 for a non-existent subprocess. Forks do occasionally fail when resources
4585 run out. I reviewed all the other calls to fork(); they all seem to check
4588 PH/20 When checking for unexpected SMTP input at connect time (before writing
4589 the banner), Exim was not dealing correctly with a non-positive return
4590 from the read() function. If the client had disconnected by this time,
4591 the result was a log entry for a synchronization error with an empty
4592 string after "input=" when read() returned zero. If read() returned -1
4593 (an event I could not check), uninitialized data bytes were printed.
4594 There were reports of junk text (parts of files, etc) appearing after
4597 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4599 PH/22 Added support for macro redefinition, and (re)definition in between
4600 driver and ACL definitions.
4602 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4603 forgetting to use the resulting value; it was using the unexpanded value.
4605 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4606 hadn't been configured. The fix is from Juergen Kreileder, who
4607 understands it better than I do:
4609 "Here's what I see happening with three configured cyrus_sasl
4610 authenticators configured (plain, login, cram-md5):
4612 On startup auth_cyrus_sasl_init() gets called for each of these.
4613 This means three calls to sasl_listmech() without a specified mech_list.
4614 => SASL tests which mechs of all available mechs actually work
4615 => three warnings about OTP not working
4616 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4618 With the patch, sasl_listmech() also gets called three times. But now
4619 SASL's mech_list option is set to the server_mech specified in the the
4620 authenticator. Or in other words, the answer from sasl_listmech()
4621 gets limited to just the mech you're testing for (which is different
4623 => the return list contains just 'plain' or 'login', 'cram-md5' or
4624 nothing depending on the value of ob->server_mech.
4626 I've just tested the patch: Authentication still works fine,
4627 unavailable mechs specified in the exim configuration are still
4628 caught, and the auth.log warnings about OTP are gone."
4630 PH/25 When debugging is enabled, the contents of the command line are added
4631 to the debugging output, even when log_selector=+arguments is not
4634 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4635 answer is "GNU", and only if the return is "GNU/something" is the answer
4638 PH/27 $acl_verify_message is now set immediately after the failure of a
4639 verification in an ACL, and so is available in subsequent modifiers. In
4640 particular, the message can be preserved by coding like this:
4642 warn !verify = sender
4643 set acl_m0 = $acl_verify_message
4645 Previously, $acl_verify_message was set only while expanding "message"
4646 and "log_message" when a very denied access.
4648 PH/28 Modified OS/os.c-Linux with
4650 -#ifndef OS_LOAD_AVERAGE
4651 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4653 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4654 nomenclature these days.)
4656 PH/29 Installed patch from the Sieve maintainer that adds the options
4657 sieve_useraddress and sieve_subaddress to the redirect router.
4659 PH/30 In these circumstances:
4660 . Two addresses routed to the same list of hosts;
4661 . First host does not offer TLS;
4662 . First host accepts first address;
4663 . First host gives temporary error to second address;
4664 . Second host offers TLS and a TLS session is established;
4665 . Second host accepts second address.
4666 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4667 and peerdn, if requested) that were in fact used only for the second
4670 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4671 attention to any local part prefix or suffix that was matched by the
4672 router that accepted the address. It now behaves in the same way as it
4673 does for delivery: the affixes are removed from the local part unless
4674 rcpt_include_affixes is set on the transport.
4676 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4677 timeout during the DATA phase of an incoming message.
4679 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4680 applied a patch sanctioned by the Sieve maintainer.
4682 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4683 the uid or gid is negative. A case of a negative gid caused this to be
4684 noticed. The fix allows for either to be negative.
4686 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4687 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4690 PH/36 The change PH/12 above was broken. Fixed it.
4692 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4693 the grounds that routing the same address twice would always produce the
4694 same answer. This might have been true once, but it is certainly no
4695 longer true now. Routing a child address may depend on the previous
4696 routing that produced that child. Some complicated redirection strategies
4697 went wrong when messages had multiple recipients, and made Exim's
4698 behaviour dependent on the order in which the addresses were given.
4700 I have moved the duplicate checking until after the routing is complete.
4701 Exim scans the addresses that are assigned to local and remote
4702 transports, and removes any duplicates. This means that more work will be
4703 done, as duplicates will always all be routed, but duplicates are
4704 presumably rare, so I don't expect this is of any significance.
4706 For deliveries to pipes, files, and autoreplies, the duplicate checking
4707 still happens during the routing process, since they are not going to be
4710 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4711 It corrects a timeout issue with spamd. This is Ian's comment: "The
4712 background is that sometimes spamd either never reads data from a
4713 connection it has accepted, or it never writes response data. The exiscan
4714 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4715 blindly assumes that writes won't block so it may never time out."
4717 PH/39 Allow G after quota size as well as K and M.
4719 PH/40 The value set for $authenticated_id in an authenticator may not contain
4720 binary zeroes or newlines because the value is written to log lines and
4721 to spool files. There was no check on this. Now the value is run through
4722 the string_printing() function so that such characters are converted to
4723 printable escape sequences.
4725 PH/41 $message_linecount is a new variable that contains the total number of
4726 lines in the message. Compare $body_linecount, which is the count for the
4729 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4730 bounce and delay warning messages, except in certain special cases, which
4733 (a) An SMTP error message from a remote host;
4734 (b) A message specified in a :fail: redirection;
4735 (c) A message specified in a "fail" command in a system filter;
4736 (d) A message specified in a FAIL return from the queryprogram router;
4737 (e) A message specified by the cannot_route_message router option.
4739 In these cases only, Exim does include the error details in bounce and
4740 warning messages. There are also a few cases where bland messages such
4741 as "unrouteable address" or "local delivery error" are given.
4743 PH/43 $value is now also set for the "else" part of a ${run expansion.
4745 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4746 being worked on, but at least Exim now implements the latest version to
4749 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4750 process to complete was treated as a delivery failure, a timeout while
4751 writing the message to the pipe was logged, but erroneously treated as a
4752 successful delivery. Such timeouts include transport filter timeouts. For
4753 consistency with the overall process timeout, these timeouts are now
4754 treated as errors, giving rise to delivery failures by default. However,
4755 there is now a new Boolean option for the pipe transport called
4756 timeout_defer, which, if set TRUE, converts the failures into defers for
4757 both kinds of timeout. A transport filter timeout is now identified in
4760 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4761 systems where "make" and "gmake" are different, calling "gmake" at top
4762 level broke things. I've arranged for the value of $(MAKE) to be passed
4763 from the Makefile to this script so that it can call the same version of
4767 A note about Exim versions 4.44 and 4.50
4768 ----------------------------------------
4770 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4771 changes of various kinds. As a consequence, a big documentation update was
4772 needed. This delayed the release for rather longer than seemed good, especially
4773 in the light of a couple of (minor) security issues. Therefore, the changes
4774 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4775 release. So 4.44 and 4.50 are in effect two different branches that both start
4778 I have left the 4.50 change log unchanged; it contains all the changes since
4779 4.43. The change log for 4.44 is below; many of its items are identical to
4780 those for 4.50. This seems to be the most sensible way to preserve the
4781 historical information.
4787 1. Minor wording change to the doc/README.SIEVE file.
4789 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4790 computation of the current number of files was incorrect.
4792 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4793 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4796 4. Give more explanation in the error message when the command for a transport
4797 filter fails to execute.
4799 5. There are several places where Exim runs a non-Exim command in a
4800 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4801 was being done only for the command run by the queryprogram router. It is
4802 now done for all such subprocesses. The other cases are: ${run, transport
4803 filters, and the commands run by the lmtp and pipe transports.
4805 6. Added CONFIGURE_GROUP build-time option.
4807 7. Some older OS have a limit of 256 on the maximum number of file
4808 descriptors. Exim was using setrlimit() to set 1000 as a large value
4809 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4810 systems. I've change it so that if it can't get 1000, it tries for 256.
4812 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4813 was an oversight, and furthermore, ever since the addition of extra
4814 controls (e.g. 4.43/32), the checks on when to allow different forms of
4815 "control" were broken. There should now be diagnostics for all cases when a
4816 control that does not make sense is encountered.
4818 9. Added the /retain_sender option to "control=submission".
4820 10. $recipients is now available in the predata ACL (oversight).
4822 11. Tidy the search cache before the fork to do a delivery from a message
4823 received from the command line. Otherwise the child will trigger a lookup
4824 failure and thereby defer the delivery if it tries to use (for example) a
4825 cached ldap connection that the parent has called unbind on.
4827 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4828 of $address_data from the recipient verification was clobbered by the
4829 sender verification.
4831 13. The value of address_data from a sender verification is now available in
4832 $sender_address_data in subsequent conditions in the ACL statement.
4834 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4836 15. Added a new option "connect=<time>" to callout options, to set a different
4839 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4840 was its contents. (It was OK if the option was not defined at all.)
4842 17. A "Completed" log line is now written for messages that are removed from
4843 the spool by the -Mrm option.
4845 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4846 information about exactly what failed.
4848 19. Added -dd to debug only the daemon process.
4850 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4851 handles timeouts, both on the server side and network timeouts. Renamed the
4852 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4854 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4855 It is now set to "smtps".
4857 22. $host_address is now set to the target address during the checking of
4858 ignore_target_hosts.
4860 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4861 being passed; this would have caused $sender_host_name to have been used if
4862 matching the list had actually called for a host name (not very likely,
4863 since this list is usually IP addresses). A host name is now passed as
4866 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4867 code that specifies a non-restarting handler (typically sigaction() in
4868 modern systems) in an attempt to fix a rare and obscure crash bug.
4870 25. Narrowed the window for a race in the daemon that could cause it to ignore
4871 SIGCHLD signals. This is not a major problem, because they are used only to
4872 wake it up if nothing else does.
4874 26. A malformed maildirsize file could cause Exim to calculate negative values
4875 for the mailbox size or file count. Odd effects could occur as a result.
4876 The maildirsize information is now recalculated if the size or filecount
4879 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4880 support for a long time. Removed HAVE_SYS_VFS_H.
4882 28. Installed the latest version of exipick from John Jetmore.
4884 29. In an address list, if the pattern was not a regular expression, an empty
4885 subject address (from a bounce message) matched only if the pattern was an
4886 empty string. Non-empty patterns were not even tested. This was the wrong
4887 because it is perfectly reasonable to use an empty address as part of a
4888 database query. An empty address is now tested by patterns that are
4889 lookups. However, all the other forms of pattern expect the subject to
4890 contain a local part and a domain, and therefore, for them, an empty
4891 address still always fails if the pattern is not itself empty.
4893 30. Exim went into a mad DNS loop when attempting to do a callout where the
4894 host was specified on an smtp transport, and looking it up yielded more
4895 than one IP address.
4897 31. Re-factored the code for checking spool and log partition space into a
4898 function that finds that data and another that does the check. The former
4899 is then used to implement four new variables: $spool_space, $log_space,
4900 $spool_inodes, and $log_inodes.
4902 32. The RFC2047 encoding function was originally intended for short strings
4903 such as real names; it was not keeping to the 75-character limit for
4904 encoded words that the RFC imposes. It now respects the limit, and
4905 generates multiple encoded words if necessary. To be on the safe side, I
4906 have increased the buffer size for the ${rfc2047: expansion operator from
4909 33. It is now permitted to omit both strings after an "if" condition; if the
4910 condition is true, the result is "true". As before, when the second string
4911 is omitted, a false condition yields an empty string. This makes it less
4912 cumbersome to write custom ACL and router conditions.
4914 34. Failure to deliver a bounce message always caused it to be frozen, even if
4915 there was an errors_to setting on the router. The errors_to setting is now
4918 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4919 canonical form (fully expanded) before being placed in
4920 $sender_host_address.
4922 36. The table in the code that translates DNS record types into text (T_A to
4923 "A" for instance) was missing entries for NS and CNAME. It is just possible
4924 that this could have caused confusion if both these types were looked up
4925 for the same domain, because the text type is used as part of Exim's
4926 per-process caching. But the chance of anyone hitting this buglet seems
4929 37. The dnsdb lookup has been extended in a number of ways.
4931 (1) There is a new type, "zns", which walks up the domain tree until it
4932 finds some nameserver records. It should be used with care.
4934 (2) There is a new type, "mxh", which is like "mx" except that it returns
4935 just the host names, not the priorities.
4937 (3) It is now possible to give a list of domains (or IP addresses) to be
4938 looked up. The behaviour when one of the lookups defers can be
4939 controlled by a keyword.
4941 (4) It is now possible to specify the separator character for use when
4942 multiple records are returned.
4944 38. The dnslists ACL condition has been extended: it is now possible to supply
4945 a list of IP addresses and/or domains to be looked up in a particular DNS
4948 39. Added log_selector=+queue_time_overall.
4950 40. When running the queue in the test harness, wait just a tad after forking a
4951 delivery process, to get repeatability of debugging output.
4953 41. Include certificate and key file names in error message when GnuTLS fails
4954 to set them up, because the GnuTLS error message doesn't include the name
4955 of the failing file when there is a problem reading it.
4957 42. Allow both -bf and -bF in the same test run.
4959 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4961 44. The "Exiscan patch" is now merged into the mainline Exim source.
4963 45. Sometimes the final signoff response after QUIT could fail to get
4964 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4965 < 0 before doing a fflush(). This bug looks as though it goes back to the
4966 introduction of TLS in release 3.20, but "sometimes" must have been rare
4967 because the tests only now provoked it.
4969 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4970 (this can affect the format of dates).
4972 47. exim_tidydb, when checking for the continued existence of a message for
4973 which it has found a message-specific retry record, was not finding
4974 messages that were in split spool directories. Consequently, it was
4975 deleting retry records that should have stayed in existence.
4977 48. Steve fixed some bugs in eximstats.
4979 49. The SPA authentication driver was not abandoning authentication and moving
4980 on to the next authenticator when an expansion was forced to fail,
4981 contradicting the general specification for all authenticators. Instead it
4982 was generating a temporary error. It now behaves as specified.
4984 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4985 (the order specifies the preference for clients). The order is now AES256,
4986 AES128, 3DES, ARCFOUR128.
4988 51. Small patch to Sieve code - explicitly set From: when generating an
4991 52. Exim crashed if a remote delivery caused a very long error message to be
4992 recorded - for instance if somebody sent an entire SpamAssassin report back
4993 as a large number of 550 error lines. This bug was coincidentally fixed by
4994 increasing the size of one of Exim's internal buffers (big_buffer) that
4995 happened as part of the Exiscan merge. However, to be on the safe side, I
4996 have made the code more robust (and fixed the comments that describe what
4999 53. Now that there can be additional text after "Completed" in log lines (if
5000 the queue_time_overall log selector is set), a one-byte patch to exigrep
5001 was needed to allow it to recognize "Completed" as not the last thing in
5004 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5005 patch that reportedly fixes this has been added. I am not expert enough to
5006 create a test for it. This is what the patch creator wrote:
5008 "I found a little strange behaviour of ldap code when working with
5009 Windows 2003 AD Domain, where users was placed in more than one
5010 Organization Units. When I tried to give exim partial DN, the exit code
5011 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5012 But simultaneously result of request was absolutely normal ldap result,
5013 so I produce this patch..."
5015 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5016 REFERENCE, so I have modified the code to exclude the patch when that macro
5019 55. Some experimental protocols are using DNS PTR records for new purposes. The
5020 keys for these records are domain names, not reversed IP addresses. The
5021 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5022 leaves it alone. Component reversal etc. now happens only for IP addresses.
5025 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5027 57. Double the size of the debug message buffer (to 2048) so that more of very
5028 long debug lines gets shown.
5030 58. The exicyclog utility now does better if the number of log files to keep
5031 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5033 59. Two changes related to the smtp_active_hostname option:
5035 (1) $smtp_active_hostname is now available as a variable.
5036 (2) The default for smtp_banner uses $smtp_active_hostname instead
5037 of $primary_hostname.
5039 60. The host_aton() function is supposed to be passed a string that is known
5040 to be a valid IP address. However, in the case of IPv6 addresses, it was
5041 not checking this. This is a hostage to fortune. Exim now panics and dies
5042 if the condition is not met. A case was found where this could be provoked
5043 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5044 components; fortuitously, this particular loophole had already been fixed
5045 by change 4.50/55 above.
5047 If there are any other similar loopholes, the new check in host_aton()
5048 itself should stop them being exploited. The report I received stated that
5049 data on the command line could provoke the exploit when Exim was running as
5050 exim, but did not say which command line option was involved. All I could
5051 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5052 running as the user.
5055 61. There was a buffer overflow vulnerability in the SPA authentication code
5056 (which came originally from the Samba project). I have added a test to the
5057 spa_base64_to_bits() function which I hope fixes it.
5060 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5061 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5063 63. The daemon start-up calls getloadavg() while still root for those OS that
5064 need the first call to be done as root, but it missed one case: when
5065 deliver_queue_load_max is set with deliver_drop_privilege. This is
5066 necessary for the benefit of the queue runner, because there is no re-exec
5067 when deliver_drop_privilege is set.
5069 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5070 This has been fixed.
5072 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5073 was in use, was not putting the data itself into the right store pool;
5074 consequently, it could be overwritten for a subsequent message in the same
5075 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5078 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5080 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5081 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5082 the function were treating the return as a boolean value, which happened to
5083 work because 0=false and not-0=true, but is not correct code.
5085 68. The host_aton() function was not handling scoped IPv6 addresses (those
5086 with, for example, "%eth0" on the end) correctly.
5088 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5089 negated items (that is, ~something) in unsigned ints. Some compilers
5090 apparently mutter when there is no cast.
5092 70. If an address verification called from an ACL failed, and did not produce a
5093 user-specific message (i.e. there was only a "system" message), nothing was
5094 put in $acl_verify_message. In this situation, it now puts the system
5097 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5098 session; change 4.31/43 added the unwanted input to the log line - except
5099 that it did not do this in the start of session case. It now does.
5101 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5102 This is wrong and can cause the other end to generate a synchronization
5103 error if it is another Exim or anything else that does the synchronization
5104 check. A QUIT command is no longer sent after a timeout.
5106 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5107 during host lookups.
5109 74. The defer_ok option of callout verification was not working if it was used
5110 when verifying addresses in header lines, that is, for this case:
5112 verify = header_sender/callout=defer_ok
5114 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5115 those file descriptors could be used for SMTP connections. If anything
5116 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5117 could be sent to the SMTP client, causing chaos. The daemon now opens
5118 stdin, stdout, and stderr to /dev/null when it puts itself into the
5121 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5122 log by default. The user can override this with suitable Perl magic.
5124 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5125 the log message when discard triggers, was not working for the DATA ACL or
5126 for the non-SMTP ACL.
5128 78. Error message wording change in sieve.c.
5130 79. If smtp_accept_max_per_host was set, the number of connections could be
5131 restricted to fewer than expected, because the daemon was trying to set up
5132 a new connection before checking whether the processes handling previous
5133 connections had finished. The check for completed processes is now done
5134 earlier. On busy systems, this bug wouldn't be noticed because something
5135 else would have woken the daemon, and it would have reaped the completed
5138 80. If a message was submitted locally by a user whose login name contained one
5139 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5140 It caused a spool format error. I have fixed the spool reading code. A
5141 related problem was that the "from" clause in the Received: line became
5142 illegal because of the space(s). It is now covered by ${quote_local_part.
5144 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5147 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5149 83. Patch from Sieve maintainer to fix unterminated string problem in
5150 "vacation" handling.
5152 84. Some minor changes to the Linux configuration files to help with other
5153 OS variants using glibc.
5155 85. One more patch for Sieve to update vacation handling to latest spec.
5158 ----------------------------------------------------
5159 See the note above about the 4.44 and 4.50 releases.
5160 ----------------------------------------------------
5166 1. Change 4.43/35 introduced a bug that caused file counts to be
5167 incorrectly computed when quota_filecount was set in an appendfile
5170 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5171 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5174 3. Give more explanation in the error message when the command for a transport
5175 filter fails to execute.
5177 4. There are several places where Exim runs a non-Exim command in a
5178 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5179 was being done only for the command run by the queryprogram router. It is
5180 now done for all such subprocesses. The other cases are: ${run, transport
5181 filters, and the commands run by the lmtp and pipe transports.
5183 5. Some older OS have a limit of 256 on the maximum number of file
5184 descriptors. Exim was using setrlimit() to set 1000 as a large value
5185 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5186 systems. I've change it so that if it can't get 1000, it tries for 256.
5188 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5189 was an oversight, and furthermore, ever since the addition of extra
5190 controls (e.g. 4.43/32), the checks on when to allow different forms of
5191 "control" were broken. There should now be diagnostics for all cases when a
5192 control that does not make sense is encountered.
5194 7. $recipients is now available in the predata ACL (oversight).
5196 8. Tidy the search cache before the fork to do a delivery from a message
5197 received from the command line. Otherwise the child will trigger a lookup
5198 failure and thereby defer the delivery if it tries to use (for example) a
5199 cached ldap connection that the parent has called unbind on.
5201 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5202 of $address_data from the recipient verification was clobbered by the
5203 sender verification.
5205 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5206 was its contents. (It was OK if the option was not defined at all.)
5208 11. A "Completed" log line is now written for messages that are removed from
5209 the spool by the -Mrm option.
5211 12. $host_address is now set to the target address during the checking of
5212 ignore_target_hosts.
5214 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5215 being passed; this would have caused $sender_host_name to have been used if
5216 matching the list had actually called for a host name (not very likely,
5217 since this list is usually IP addresses). A host name is now passed as
5220 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5221 code that specifies a non-restarting handler (typically sigaction() in
5222 modern systems) in an attempt to fix a rare and obscure crash bug.
5224 15. Narrowed the window for a race in the daemon that could cause it to ignore
5225 SIGCHLD signals. This is not a major problem, because they are used only to
5226 wake it up if nothing else does.
5228 16. A malformed maildirsize file could cause Exim to calculate negative values
5229 for the mailbox size or file count. Odd effects could occur as a result.
5230 The maildirsize information is now recalculated if the size or filecount
5233 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5234 support for a long time. Removed HAVE_SYS_VFS_H.
5236 18. Updated exipick to current release from John Jetmore.
5238 19. Allow an empty sender to be matched against a lookup in an address list.
5239 Previously the only cases considered were a regular expression, or an
5242 20. Exim went into a mad DNS lookup loop when doing a callout where the
5243 host was specified on the transport, if the DNS lookup yielded more than
5246 21. The RFC2047 encoding function was originally intended for short strings
5247 such as real names; it was not keeping to the 75-character limit for
5248 encoded words that the RFC imposes. It now respects the limit, and
5249 generates multiple encoded words if necessary. To be on the safe side, I
5250 have increased the buffer size for the ${rfc2047: expansion operator from
5253 22. Failure to deliver a bounce message always caused it to be frozen, even if
5254 there was an errors_to setting on the router. The errors_to setting is now
5257 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5258 canonical form (fully expanded) before being placed in
5259 $sender_host_address.
5261 24. Updated eximstats to version 1.33
5263 25. Include certificate and key file names in error message when GnuTLS fails
5264 to set them up, because the GnuTLS error message doesn't include the name
5265 of the failing file when there is a problem reading it.
5267 26. Expand error message when OpenSSL has problems setting up cert/key files.
5270 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5271 (this can affect the format of dates).
5273 28. exim_tidydb, when checking for the continued existence of a message for
5274 which it has found a message-specific retry record, was not finding
5275 messages that were in split spool directories. Consequently, it was
5276 deleting retry records that should have stayed in existence.
5278 29. eximstats updated to version 1.35
5279 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5280 1.35 - bugfix such that pie charts by volume are generated correctly
5282 30. The SPA authentication driver was not abandoning authentication and moving
5283 on to the next authenticator when an expansion was forced to fail,
5284 contradicting the general specification for all authenticators. Instead it
5285 was generating a temporary error. It now behaves as specified.
5287 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5288 (the order specifies the preference for clients). The order is now AES256,
5289 AES128, 3DES, ARCFOUR128.
5291 31. Small patch to Sieve code - explicitly set From: when generating an
5294 32. Exim crashed if a remote delivery caused a very long error message to be
5295 recorded - for instance if somebody sent an entire SpamAssassin report back
5296 as a large number of 550 error lines. This bug was coincidentally fixed by
5297 increasing the size of one of Exim's internal buffers (big_buffer) that
5298 happened as part of the Exiscan merge. However, to be on the safe side, I
5299 have made the code more robust (and fixed the comments that describe what
5302 33. Some experimental protocols are using DNS PTR records for new purposes. The
5303 keys for these records are domain names, not reversed IP addresses. The
5304 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5305 leaves it alone. Component reversal etc. now happens only for IP addresses.
5308 34. The host_aton() function is supposed to be passed a string that is known
5309 to be a valid IP address. However, in the case of IPv6 addresses, it was
5310 not checking this. This is a hostage to fortune. Exim now panics and dies
5311 if the condition is not met. A case was found where this could be provoked
5312 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5313 components; fortuitously, this particular loophole had already been fixed
5314 by change 4.50/55 or 4.44/33 above.
5316 If there are any other similar loopholes, the new check in host_aton()
5317 itself should stop them being exploited. The report I received stated that
5318 data on the command line could provoke the exploit when Exim was running as
5319 exim, but did not say which command line option was involved. All I could
5320 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5321 running as the user.
5324 35. There was a buffer overflow vulnerability in the SPA authentication code
5325 (which came originally from the Samba project). I have added a test to the
5326 spa_base64_to_bits() function which I hope fixes it.
5329 36. The daemon start-up calls getloadavg() while still root for those OS that
5330 need the first call to be done as root, but it missed one case: when
5331 deliver_queue_load_max is set with deliver_drop_privilege. This is
5332 necessary for the benefit of the queue runner, because there is no re-exec
5333 when deliver_drop_privilege is set.
5335 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5336 was in use, was not putting the data itself into the right store pool;
5337 consequently, it could be overwritten for a subsequent message in the same
5338 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5341 38. Sometimes the final signoff response after QUIT could fail to get
5342 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5343 < 0 before doing a fflush(). This bug looks as though it goes back to the
5344 introduction of TLS in release 3.20, but "sometimes" must have been rare
5345 because the tests only now provoked it.
5351 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5352 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5353 Now it returns an integer. A number of calls were still expecting a T/F
5354 return. Fortuitously, in all cases, the tests worked in OK situations,
5355 which is the norm. However, things would have gone wrong on any write
5356 failures on the smtp file descriptor. This function is used when sending
5357 messages over SMTP and also when doing verify callouts.
5359 2. When Exim is called to do synchronous delivery of a locally submitted
5360 message (the -odf or -odi options), it no longer closes stderr before doing
5363 3. Implemented the mua_wrapper option.
5365 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5367 5. Implemented the functions header_remove(), header_testname(),
5368 header_add_at_position(), and receive_remove_recipient(), and exported them
5371 6. If an ACL "warn" statement specified the addition of headers, Exim already
5372 inserted X-ACL-Warn: at the start if there was no header name. However, it
5373 was not making this test for the second and subsequent header lines if
5374 there were newlines in the string. This meant that an invalid header could
5375 be inserted if Exim was badly configured.
5377 7. Allow an ACL "warn" statement to add header lines at the start or after all
5378 the Received: headers, as well as at the end.
5380 8. Added the rcpt_4xx retry error code.
5382 9. Added postmaster_mailfrom=xxx to callout verification option.
5384 10. Added mailfrom=xxxx to the callout verification option, for verify=
5387 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5388 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5390 12. Inserted some casts to stop certain compilers warning when using pointer
5391 differences as field lengths or precisions in printf-type calls (mostly
5392 affecting debugging statements).
5394 13. Added optional readline() support for -be (dynamically loaded).
5396 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5397 same clock tick as a message's arrival, so that its received time was the
5398 same as the "first fail" time on the retry record, and that message
5399 remained on the queue past the ultimate address timeout, every queue runner
5400 would try a delivery (because it was past the ultimate address timeout) but
5401 after another failure, the ultimate address timeout, which should have then
5402 bounced the address, did not kick in. This was a "< instead of <=" error;
5403 in most cases the first failure would have been in the next clock tick
5404 after the received time, and all would be well.
5406 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5407 being recognized when the domain list was tested by the match_domain
5408 condition in an expansion string.
5410 16. Added the ${str2b64: operator.
5412 17. Exim was always calling setrlimit() to set a large limit for the number of
5413 processes, without checking whether the existing limit was already
5414 adequate. (It did check for the limit on file descriptors.) Furthermore,
5415 errors from getrlimit() and setrlimit() were being ignored. Now they are
5416 logged to the main and panic logs, but Exim does carry on, to try to do its
5417 job under whatever limits there are.
5419 18. Imported PCRE 5.0.
5421 19. Trivial typo in log message " temporarily refused connection" (the leading
5424 20. If the log selector return_path_on_delivery was set and an address was
5425 redirected to /dev/null, the delivery process crashed because it assumed
5426 that a return path would always be set for a "successful" delivery. In this
5427 case, the whole delivery is bypassed as an optimization, and therefore no
5430 21. Internal re-arrangement: the function for sending a challenge and reading
5431 a response while authentication was assuming a zero-terminated challenge
5432 string. It's now changed to take a pointer and a length, to allow for
5433 binary data in such strings.
5435 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5437 23. Exim was not respecting finduser_retries when seeking the login of the
5438 uid under which it was called; it was always trying 10 times. (The default
5439 setting of finduser_retries is zero.) Also, it was sleeping after the final
5440 failure, which is pointless.
5442 24. Implemented tls_on_connect_ports.
5444 25. Implemented acl_smtp_predata.
5446 26. If the domain in control=submission is set empty, Exim assumes that the
5447 authenticated id is a complete email address when it generates From: or
5448 Sender: header lines.
5450 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5451 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5452 chown and chgrp in /bin and hostname in /usr/bin.
5454 28. Exim was keeping the "process log" file open after each use, just as it
5455 does for the main log. This opens the possibility of it remaining open for
5456 long periods when the USR1 signal hits a daemon. Occasional processlog
5457 errors were reported, that could have been caused by this. Anyway, it seems
5458 much more sensible not to leave this file open at all, so that is what now
5461 29. The long-running daemon process does not normally write to the log once it
5462 has entered its main loop, and it closes the log before doing so. This is
5463 so that log files can straightforwardly be renamed and moved. However,
5464 there are a couple of unusual error situations where the daemon does write
5465 log entries, and I had neglected to close the log afterwards.
5467 30. The text of an SMTP error response that was received during a remote
5468 delivery was being truncated at 512 bytes. This is too short for some of
5469 the long messages that one sometimes sees. I've increased the limit to
5472 31. It is now possible to make retry rules that apply only when a message has a
5473 specific sender, in particular, an empty sender.
5475 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5476 it possible to be selective about when SMTP synchronization is enforced.
5478 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5480 32. Implemented hosts_connection_nolog.
5482 33. Added an ACL for QUIT.
5484 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5487 35. Added mailbox_size and mailbox_filecount to appendfile.
5489 36. Added control = no_multiline_responses to ACLs.
5491 37. There was a bug in the logic of the code that waits for the clock to tick
5492 in the case where the clock went backwards by a substantial amount such
5493 that the microsecond fraction of "now" was more than the microsecond
5494 fraction of "then" (but the whole seconds number was less).
5496 38. Added support for the libradius Radius client library this is found on
5497 FreeBSD (previously only the radiusclient library was supported).
5503 1. When certain lookups returned multiple values in the form name=value, the
5504 quoting of the values was not always being done properly. Specifically:
5505 (a) If the value started with a double quote, but contained no whitespace,
5507 (b) If the value contained whitespace other than a space character (i.e.
5508 tabs or newlines or carriage returns) it was not quoted.
5509 This fix has been applied to the mysql and pgsql lookups by writing a
5510 separate quoting function and calling it from the lookup code. The fix
5511 should probably also be applied to nisplus, ibase and oracle lookups, but
5512 since I cannot test any of those, I have not disturbed their existing code.
5514 2. A hit in the callout cache for a specific address caused a log line with no
5515 reason for rejecting RCPT. Now it says "Previous (cached) callout
5516 verification failure".
5518 3. There was an off-by-one bug in the queryprogram router. An over-long
5519 return line was truncated at 256 instead of 255 characters, thereby
5520 overflowing its buffer with the terminating zero. As well as fixing this, I
5521 have increased the buffer size to 1024 (and made a note to document this).
5523 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5524 when Exim is waiting for an SMTP response from a remote server, Exim
5525 restarts its select() call on the socket, thereby resetting its timeout.
5526 This is not a problem when such interrupts are rare. Somebody set up a cron
5527 job to run exiwhat every 2 minutes, which is less than the normal select()
5528 timeout (5 or 10 minutes). This meant that the select() timeout never
5529 kicked in because it was always reset. I have fixed this by comparing the
5530 time when an interrupt arrives with the time at the start of the first call
5531 to select(). If more time than the timeout has elapsed, the interrupt is
5532 treated as a timeout.
5534 5. Some internal re-factoring in preparation for the addition of Sieve
5535 extensions (by MH). In particular, the "personal" test is moved to a
5536 separate function, and given an option for scanning Cc: and Bcc: (which is
5537 not set for Exim filters).
5539 6. When Exim created an email address using the login of the caller as the
5540 local part (e.g. when creating a From: or Sender: header line), it was not
5541 quoting the local part when it contained special characters such as @.
5543 7. Installed new OpenBSD configuration files.
5545 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5546 try to make them clearer.
5548 9. Callout options, other than the timeout value, were being ignored when
5549 verifying sender addresses in header lines. For example, when using
5551 verify = header_sender/callout=no_cache
5553 the cache was (incorrectly) being used.
5555 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5556 only the Cygwin environment.
5558 11. When return_path_on_delivery was set as a log selector, if different remote
5559 addresses in the same message used different return paths and parallel
5560 remote delivery occurred, the wrong values would sometimes be logged.
5561 (Whenever a remote delivery process finished, the return path value from
5562 the most recently started remote delivery process was logged.)
5564 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5565 lines when AUTH and/or TLS are in use. This is the "received protocol"
5566 field. Exim used to use "asmtp" for authenticated SMTP, without any
5567 indication (in the protocol name) for TLS use. Now it follows the RFC and
5568 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5569 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5570 names appear in log lines as well as in Received: header lines.
5572 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5573 extensions, and comparison tests, and to fix some bugs.
5575 14. Changes to the "personal" filter test:
5577 (1) The test was buggy in that it was just doing the equivalent of
5578 "contains" tests on header lines. For example, if a user's address was
5579 anne@some.where, the "personal" test would incorrectly be true for
5581 To: susanne@some.where
5583 This test is now done by extracting each address from the header in turn,
5584 and checking the entire address. Other tests that are part of "personal"
5585 are now done using regular expressions (for example, to check local parts
5586 of addresses in From: header lines).
5588 (2) The list of non-personal local parts in From: addresses has been
5589 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5590 taken from the Sieve specification recommendations.
5592 (3) If the message contains any header line starting with "List-" it is
5593 treated as non-personal.
5595 (4) The test for "circular" in the Subject: header line has been removed
5596 because it now seems ill-conceived.
5598 15. Minor typos in src/EDITME comments corrected.
5600 16. Installed latest exipick from John Jetmore.
5602 17. If headers_add on a router specified a text string that was too long for
5603 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5604 of string_sprintf() is now avoided.
5606 18. $message_body_size was not set (it was always zero) when running the DATA
5607 ACL and the local_scan() function.
5609 19. For the "mail" command in an Exim filter, no default was being set for
5610 the once_repeat time, causing a random time value to be used if "once" was
5611 specified. (If the value happened to be <= 0, no repeat happened.) The
5612 default is now 0s, meaning "never repeat". The "vacation" command was OK
5613 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5614 (I found it when inspecting the code).
5616 20. There is now an overall timeout for performing a callout verification. It
5617 defaults to 4 times the callout timeout, which applies to individual SMTP
5618 commands during the callout. The overall timeout applies when there is more
5619 than one host that can be tried. The timeout is checked before trying the
5620 next host. This prevents very long delays if there are a large number of
5621 hosts and all are timing out (e.g. when the network connections are timing
5622 out). The value of the overall timeout can be changed by specifying an
5623 additional sub-option for "callout", called "maxwait". For example:
5625 verify = sender/callout=5s,maxwait=20s
5627 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5628 to the end before writing, but this should make it even safer).
5630 22. Exim was forgetting that it had advertised PIPELINING for the second and
5631 subsequent messages on an SMTP connection. It was also not resetting its
5632 memory on STARTTLS and an internal HELO.
5634 23. When Exim logs an SMTP synchronization error within a session, it now
5635 records whether PIPELINING has been advertised or not.
5637 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5638 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5639 rather than long int.
5641 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5643 26. Added the never_mail option to autoreply.
5649 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5650 crash if the getsockname() call failed; this can happen if a connection is
5651 closed very soon after it is established. The problem was simply in the
5652 order in which certain operations were done, causing Exim to try to write
5653 to the SMTP stream before it had set up the file descriptor. The bug has
5654 been fixed by making things happen in the correct order.
5660 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5661 before the connection was closed, thus losing the rejection response.
5663 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5664 some early Solaris releases, but causes trouble in current releases where
5665 socklen_t is defined.
5667 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5670 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5673 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5674 that releasing the top of it at the end releases what was used for sub-
5675 expansions (unless the block got too big). However, discard this block if
5676 the first thing is a variable or header, so that we can use its block when
5677 it is dynamic (useful for very large $message_headers, for example).
5679 6. Lookups now cache *every* query, not just the most recent. A new, separate
5680 store pool is used for this. It can be recovered when all lookup caches are
5681 flushed. Lookups now release memory at the end of their result strings.
5682 This has involved some general refactoring of the lookup sources.
5684 7. Some code has been added to the store_xxx() functions to reduce the amount
5685 of flapping under certain conditions.
5687 8. log_incoming_interface used to affect only the <= reception log lines. Now
5688 it causes the local interface and port to be added to several more SMTP log
5689 lines, for example "SMTP connection from", and rejection lines.
5691 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5693 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5695 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5696 could be overwritten at the end of the current message (or the start of a
5697 new message if it was set in a HELO ACL). The value is now preserved for
5698 the duration of the SMTP connection.
5700 12. If a transport had a headers_rewrite setting, and a matching header line
5701 contained an unqualified address, that address was qualified, even if it
5702 did not match any rewriting rules. The underlying bug was that the values
5703 of the flags that permit the existence of unqualified sender and recipient
5704 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5705 non-local messages, and by -bnq for local messages) were not being
5706 preserved with the message after it was received.
5708 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5709 "next input=" as part of the text comprising the host identity instead of
5710 the correct text. The code was using the same buffer for two different
5711 strings. However, depending on which order the printing function evaluated
5712 its arguments, the bug did not always show up. Under Linux, for example, my
5713 test suite worked just fine.
5715 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5716 doesn't work with some older versions of Perl. It has been changed to "my",
5717 which in any case is probably the better facility to use.
5719 15. A really picky compiler found some instances of statements for creating
5720 error messages that either had too many or two few arguments for the format
5723 16. The size of the buffer for calls to the DNS resolver has been increased
5724 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5725 for addresses that have a lot of PTR records. This alleviates a problem; it
5726 does not fully solve it.
5728 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5729 buffer now truncates the list and logs the incident, which is the same
5730 action as happens when Exim is looking up a host name and its aliases.
5731 Previously in this situation something unpredictable would happen;
5732 sometimes it was "internal error: store_reset failed".
5734 18. If a server dropped the connection unexpectedly when an Exim client was
5735 using GnuTLS and trying to read a response, the client delivery process
5736 crashed while trying to generate an error log message.
5738 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5739 string, for example:
5741 warn message = H1: something\nH2: something
5743 the text was added as a single header line from Exim's point of view
5744 though it ended up OK in the delivered message. However, searching for the
5745 second and subsequent header lines using $h_h2: did not work. This has been
5746 fixed. Similarly, if a system filter added multiple headers in this way,
5747 the routers could not see them.
5749 20. Expanded the error message when iplsearch is called with an invalid key to
5750 suggest using net-iplsearch in a host list.
5752 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5753 ACLs are no longer actually imposed (and a message to that effect is
5756 22. If a "gecos" field in a passwd entry contained escaped characters, in
5757 particular, if it contained a \" sequence, Exim got it wrong when building
5758 a From: or a Sender: header from that name. A second bug also caused
5759 incorrect handling when an unquoted " was present following a character
5760 that needed quoting.
5762 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5763 was not being matched caselessly.
5765 24. Arranged for all hyphens in the exim.8 source to be escaped with
5768 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5769 properly. Recipient callout cache records were still being keyed to include
5770 the sender, even when use_sender was set false. This led to far more
5771 callouts that were necessary. The sender is no longer included in the key
5772 when use_sender is false.
5774 26. Added "control = submission" modifier to ACLs.
5776 27. Added the ${base62d: operator to decode base 62 numbers.
5778 28. dnsdb lookups can now access SRV records.
5780 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5781 the configuration file.
5783 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5784 (-v) mode. This makes the output for a verbose queue run more intelligible.
5786 31. Added a use_postmaster feature to recipient callouts.
5788 32. Added the $body_zerocount variable, containing the number of binary zero
5789 bytes in the message body.
5791 33. The time of last modification of the "new" subdirectory is now used as the
5792 "mailbox time last read" when there is a quota error for a maildir
5795 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5797 35. Added +ignore_unknown as a special item in host lists.
5799 36. Code for decoding IPv6 addresses in host lists is now included, even if
5800 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5801 address was recognized as an IP address, but was then not correctly decoded
5802 into binary, causing unexpected and incorrect effects when compared with
5809 1. Very minor rewording of debugging text in manualroute to say "list of
5810 hosts" instead of "hostlist".
5812 2. If verify=header_syntax was set, and a header line with an unqualified
5813 address (no domain) and a large number of spaces between the end of the
5814 name and the colon was received, the reception process suffered a buffer
5815 overflow, and (when I tested it) crashed. This was caused by some obsolete
5816 code that should have been removed. The fix is to remove it!
5818 3. When running in the test harness, delay a bit after writing a bounce
5819 message to get a bit more predictability in the log output.
5821 4. Added a call to search_tidyup() just before forking a reception process. In
5822 theory, someone could use a lookup in the expansion of smtp_accept_max_
5823 per_host which, without the tidyup, could leave open a database connection.
5825 5. Added the variables $recipient_data and $sender_data which get set from a
5826 lookup success in an ACL "recipients" or "senders" condition, or a router
5827 "senders" option, similar to $domain_data and $local_part_data.
5829 6. Moved the writing of debug_print from before to after the "senders" test
5832 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5833 problems for message scanning, either using a data ACL, or using
5834 local_scan() because the Received: header was not generated till after they
5835 were called (in order to set the time as the time of reception completion).
5836 I have revised the way this works. The header is now generated after the
5837 body is received, but before the ACL or local_scan() are called. After they
5838 are run, the timestamp in the header is updated.
5844 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5845 before starting a queue runner without re-exec. This happened only when
5846 deliver_drop_privilege was set or when the Exim user was set to root. The
5847 effect of the bug was that timeouts during subsequent deliveries caused
5848 crashes instead of being properly handled. The handler is now left at its
5849 default (and expected) setting.
5851 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5852 message, again when deliver_drop_privilege is set or Exim is run as root.
5853 The bug described in (1) was not present in this case, but the tidying up
5854 of the other signals was missing. I have made the two cases consistent.
5856 3. The ignore_target_hosts setting on a manualroute router was being ignored
5857 for hosts that were looked up using the /MX notation.
5859 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5862 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5863 operated on the sender address. After changing the $sender_address to <>
5864 for the sender address verify, Exim was re-instated it as the original
5865 (before rewriting) address, but remembering that it had rewritten it, so it
5866 wasn't rewriting it again. This bug also had the effect of breaking the
5867 sender address verification caching when the sender address was rewritten.
5869 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5870 This has been changed so that if the ip literal address matches
5871 ignore_target_hosts, the router declines.
5873 7. Added expansion conditions match_domain, match_address, and match_local_
5874 part (NOT match_host).
5876 8. The placeholder for the Received: header didn't have a length field set.
5878 9. Added code to Exim itself and to exim_lock to test for a specific race
5879 condition that could lead to file corruption when using MBX delivery. The
5880 issue is with the lockfile that is created in /tmp. If this file is removed
5881 after a process has opened it but before that process has acquired a lock,
5882 there is the potential for a second process to recreate the file and also
5883 acquire a lock. This could lead to two Exim processes writing to the file
5884 at the same time. The added code performs the same test as UW imapd; it
5885 checks after acquiring the lock that its file descriptor still refers to
5886 the same named file.
5888 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5889 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5892 11. Added the smtp_active_hostname option. If used, this will typically be made
5893 to depend on the incoming interface address. Because $interface_address is
5894 not set up until the daemon has forked a reception process, error responses
5895 that can happen earlier (such as "too many connections") no longer contain
5898 12. If an expansion in a condition on a "warn" statement fails because a lookup
5899 defers, the "warn" statement is abandoned, and the next ACL statement is
5900 processed. Previously this caused the whole ACL to be aborted.
5902 13. Added the iplsearch lookup type.
5904 14. Added ident_timeout as a log selector.
5906 15. Added tls_certificate_verified as a log selector.
5908 16. Added a global option tls_require_ciphers (compare the smtp transport
5909 option of the same name). This controls incoming TLS connections.
5911 17. I finally figured out how to make tls_require_ciphers do a similar thing
5912 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5913 before starting the TLS session.
5915 18. Tabs are now shown as \t in -bP output.
5917 19. If the log selector return_path_on_delivery was set, Exim crashed when
5918 bouncing a message because it had too many Received: header lines.
5920 20. If two routers both had headers_remove settings, and the first one included
5921 a superfluous trailing colon, the final name in the first list and the
5922 first name in the second list were incorrectly joined into one item (with a
5923 colon in the middle).
5929 1. Added -C and -D options to the exinext utility, mainly to make it easier
5930 to include in the automated testing, but these could be helpful when
5931 multiple configurations are in use.
5933 2. The exinext utility was not formatting the output nicely when there was
5934 an alternate port involved in the retry record key, nor when there was a
5935 message id as well (for retries that were specific to a specific message
5936 and a specific host). It was also confused by IPv6 addresses, because of
5937 the additional colons they contain. I have fixed the IPv4 problem, and
5938 patched it up to do a reasonable job for IPv6.
5940 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5941 delivery, the log line now contains "pipelined" if PIPELINING was used.
5943 4. An SMTP transport process used to panic and die if the bind() call to set
5944 an explicit outgoing interface failed. This has been changed; it is now
5945 treated in the same way as a connect() failure.
5947 5. A reference to $sender_host_name in the part of a conditional expansion
5948 that was being skipped was still causing a DNS lookup. This no longer
5951 6. The def: expansion condition was not recognizing references to header lines
5952 that used bh_ and bheader_.
5954 7. Added the _cache feature to named lists.
5956 8. The code for checking quota_filecount in the appendfile transport was
5957 allowing one more file than it should have been.
5959 9. For compatibility with Sendmail, the command line option
5967 and sets the incoming protocol and host name (for trusted callers). The
5968 host name and its colon can be omitted when only the protocol is to be set.
5969 Note the Exim already has two private options, -pd and -ps, that refer to
5970 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5971 "s", but I don't think that's a major issue.
5973 10. A number of refactoring changes to the code, none of which should affect
5976 (a) The number of logging options was getting close to filling up the
5977 32-bit word that was used as a bit map. I have split them into two classes:
5978 those that are passed in the argument to log_write(), and those that are
5979 only ever tested independently outside of that function. These are now in
5980 separate 32-bit words, so there is plenty of room for expansion again.
5981 There is no change in the user interface or the logging behaviour.
5983 (b) When building, for example, log lines, the code previously used a
5984 macro that called string_cat() twice, in order to add two strings. This is
5985 not really sufficiently general. Furthermore, there was one instance where
5986 it was actually wrong because one of the argument was used twice, and in
5987 one call a function was used. (As it happened, calling the function twice
5988 did not affect the overall behaviour.) The macro has been replaced by a
5989 function that can join an arbitrary number of extra strings onto a growing
5992 (c) The code for expansion conditions now uses a table and a binary chop
5993 instead of a serial search (which was left over from when there were very
5994 few conditions). Also, it now recognizes conditions like "pam" even when
5995 the relevant support is not compiled in: a suitably worded error message is
5996 given if an attempt is made to use such a condition.
5998 11. Added ${time_interval:xxxxx}.
6000 12. A bug was causing one of the ddress fields not to be passed back correctly
6001 from remote delivery subprocesses. The field in question was not being
6002 subsequently used, so this caused to problems in practice.
6004 13. Added new log selectors queue_time and deliver_time.
6006 14. Might have fixed a bug in maildirsizefile handling that threw up
6007 "unexpected character" debug warnings, and recalculated the data
6008 unnecessarily. In any case, I expanded the warning message to give more
6011 15. Added the message "Restricted characters in address" to the statements in
6012 the default ACL that block characters like @ and % in local parts.
6014 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6015 Three changes have been made:
6017 (a) There was a serious bug; a negative response to MAIL caused the whole
6018 recipient domain to be cached as invalid, thereby blocking all messages
6019 to all local parts at the same domain, from all senders. This bug has
6020 been fixed. The domain is no longer cached after a negative response to
6021 MAIL if the sender used is not empty.
6023 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6026 (c) A new callout option, "use_sender" has been added for people who want
6027 the modified behaviour.
6033 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6036 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6037 indeed breaks things for older releases.
6039 3. Added additional logging to the case where there is a problem reading data
6040 from a filter that is running in a subprocess using a pipe, in order to
6041 try to track down a specific problem.
6043 4. Testing facility fudge: when running in the test harness and attempting
6044 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6045 getting "No route to host". Convert this to a timeout.
6047 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6050 6. Some OS don't have socklen_t but use size_t instead. This affects the
6051 fifth argument of getsockopt() amongst other things. This is now
6052 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6053 can be set for individual OS. I have set it for SunOS5, OSF1, and
6054 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6055 some earlier ones do not.
6057 7. Change 4.30/15 was not doing the test caselessly.
6059 8. The standard form for an IPv6 address literal was being rejected by address
6060 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6061 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6062 this, as well as the form without the "IPv6" on the front (but only when
6063 address literals are enabled, of course).
6065 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6067 10. Exim crashed if a message with an empty sender address specified by -f
6068 encountered a router with an errors_to setting. This could be provoked only
6069 by a command such as
6073 where an empty string was supplied; "<>" did not hit this bug.
6075 11. Installed PCRE release 4.5.
6077 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6078 remained set. It is now erased.
6080 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6081 times from message ids (which are base 36 rather than the normal 62).
6083 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6084 were being counted as actual protocol errors, and logged if the log
6085 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6086 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6087 and DATA following a set of rejected RCPTs do not count as protocol errors.
6088 In other words, Exim assumes they were pipelined, though this may not
6089 actually be the case. Of course, in all cases the client gets an
6090 appropriate error code.
6092 15. If a lookup fails in an ACL condition, a message about the failure may
6093 be available; it is used if testing the ACL cannot continue, because most
6094 such messages specify what the cause of the deferral is. However, some
6095 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6096 that caused an old message to be retained and used if a later statement
6097 caused a defer, replacing the real cause of the deferral.
6099 16. If an IP address had so many PTR records that the DNS lookup buffer
6100 was not large enough to hold them, Exim could crash while trying to process
6101 the truncated data. It now detects and logs this case.
6103 17. Further to 4.21/58, another change has been made: if (and only if) the
6104 first line of a message (the first header line) ends with CRLF, a bare LF
6105 in a subsequent header line has a space inserted after it, so as not to
6106 terminate the header.
6108 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6109 unnecessarily, used atoi() instead of strtol(), and didn't check the
6110 termination when getting file sizes from file names by regex.
6112 19. Completely re-implemented the support for maildirsize files, in the light
6113 of a number of problems with the previous contributed implementation
6114 (4.30/29). In particular:
6116 . If the quota is zero, the maildirsize file is maintained, but no quota is
6119 . If the maildir directory does not exist, it is created before any attempt
6120 to write a maildirsize file.
6122 . The quota value in the file is just a cache; if the quota is changed in
6123 the transport, the new value overrides.
6125 . A regular expression is available for excluding directories from the
6128 20. The autoreply transport checks the characters in options that define the
6129 message's headers; it allows continued headers, but it was checking with
6130 isspace() after an embedded newline instead of explicitly looking for a
6133 21. If all the "regular" hosts to which an address was routed had passed their
6134 expiry times, and had not reached their retry times, the address was
6135 bounced, even if fallback hosts were defined. Now Exim should go on to try
6138 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6139 equivalent code in the SMTP transport. Some hosts send humungous responses
6140 to HELO/EHLO, more than 1024 it seems.
6142 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6143 gives compiler warnings in some environments. I've now done it "properly",
6146 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6147 (because of problems with the built-in one) was declared to return uschar *
6148 instead of char *, causing compiler failure.
6150 25. Fixed a file descriptor leak when processing alias/forward files.
6152 26. Fixed a minor format string issue in dbfn.c.
6154 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6156 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6157 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6160 29. When a sender address is verified, it is cached, to save repeating the test
6161 when there is more than one recipient in a message. However, when the
6162 verification involves a callout, it is possible for different callout
6163 options to be set for different recipients. It is too complicated to keep
6164 track of this in the cache, so now Exim always runs a verification when a
6165 callout is required, relying on the callout cache for the optimization.
6166 The overhead is duplication of the address routing, but this should not be
6169 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6170 to be verified with callout=postmaster, and the main callout worked but the
6171 postmaster check failed, the verification correctly failed. However, if a
6172 subsequent RCPT command asked for sender verification *without* the
6173 postmaster check, incorrect caching caused this verification also to fail,
6176 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6177 it was not caching the DNS options (qualify_single, search_parents) that
6178 were used when the lookup failed. A subsequent lookup with different
6179 options therefore always gave the same answer, though there were cases
6180 where it should not have. (Example: a "domains = !$mx_any" option on a
6181 dnslookup router: the "domains" option is always processed without any
6182 widening, but the router might have qualify_single set.) Now Exim uses the
6183 cached value only when the same options are set.
6185 32. Added John Jetmore's "exipick" utility to the distribution.
6187 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6188 than a timeout (e.g. a certificate is required, and is not provided), an
6189 Exim server now closes the connection immediately. Previously it waited for
6190 the client to close - but if the client is SSL, it seems that they each
6191 wait for each other, leading to a delay before one of them times out.
6193 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6194 maintained 0.8.x compatibility because I don't think many are using it, and
6195 it is clearly obsolete.
6197 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6200 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6201 client certificate was expired. A simple patch fixes this, though I don't
6202 understand the full logic of why the verify callback is called multiple
6205 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6206 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6207 which causes problems with some clients (such as the Certicom SSL Plus
6208 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6209 disables the coutermeasure allowing Eudora to connect."
6211 38. Exim was not checking that a write() to a log file succeeded. This could
6212 lead to Bad Things if a log got too big, in particular if it hit a file
6213 size limit. Exim now panics and dies if it cannot write to a log file, just
6214 as it does if it cannot open a log file.
6216 39. Modified OS/Makefile-Linux so that it now contains
6218 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6220 The two -D definitions ensure that Exim is compiled with large file
6221 support, which makes it possible to handle log files that are bigger than
6224 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6225 instance) a domain was checked against a named list that involved a lookup,
6226 causing $domain_data to be set, then another domain was checked against the
6227 same list, then the first domain was re-checked, the value of $domain_data
6228 after the final check could be wrong. In particular, if the second check
6229 failed, it could be set empty. This bug probably also applied to
6232 41. The strip_trailing_dot option was not being applied to the address given
6233 with the -f command-line option.
6235 42. The code for reading a message's header from the spool was incrementing
6236 $received_count, but never initializing it. This meant that the value was
6237 incorrect (doubled) while delivering a message in the same process in which
6238 it was received. In the most common configuration of Exim, this never
6239 happens - a fresh exec is done - but it can happen when
6240 deliver_drop_privilege is set.
6242 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6243 it now includes up to 150 characters of the unexpected data in the log
6246 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6247 and building data strings. The size of both of these buffers was 10 000
6248 bytes - far larger than anybody would *ever* want, thought I. Needless to
6249 say, somebody hit the limit. I have increased the maximum line length to
6250 20 000 and the maximum data length of concatenated lines to 100 000. I have
6251 also fixed two bugs, because there was no checking on these buffers. Tsk,
6252 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6253 buffer is too small.
6255 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6256 lsearch lookups. Now it does.
6258 46. When parsing a route_list item in a manualroute router, a fixed-length
6259 buffer was used for the list of hosts. I made this 1024 bytes long,
6260 thinking that nobody would ever have a list of hosts that long. Wrong.
6261 Somebody had a whole pile of complicated expansion conditions, and the
6262 string was silently truncated, leading to an expansion error. It turns out
6263 that it is easier to change to an unlimited length (owing to other changes
6264 that have happened since this code was originally written) than to build
6265 structure for giving a limitation error. The length of the item that
6266 expands into the list of hosts is now unlimited.
6268 47. The lsearch lookup could not handle data where the length of text line was
6269 more than 4095 characters. Such lines were truncated, leading to shortened
6270 data being returned. It should now handle lines of any length.
6272 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6273 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6276 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6277 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6278 no longer happens in comment lines. A list of replacements is now placed
6279 at the head of all of the source files, except those whose only change is
6280 to replace PERL_COMMAND in the very first #! line.
6282 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6283 messages on the queue, with a bottom-up merge sort, using code contributed
6284 by Michael Haardt. This should make operations like -bp somewhat faster on
6285 large queues. It won't affect queue runners, except when queue_run_in_order
6288 51. Installed eximstats 1.31 in the distribution.
6290 52. Added support for SRV lookups to the dnslookup router.
6292 53. If an ACL referred to $message_body or $message_body_end, the value was not
6293 reset for any messages that followed in the same SMTP session.
6295 54. The store-handling optimization for building very long strings was not
6296 differentiating between the different store pools. I don't think this
6297 actually made any difference in practice, but I've tidied it.
6299 55. While running the routers to verify a sender address, $sender_address
6300 was still set to the sender address. This is wrong, because when routing to
6301 send a bounce to the sender, it would be empty. Therefore, I have changed
6302 it so that, while verifying a sender address, $sender_address is set to <>.
6303 (There is no change to what happens when verifying a recipient address.)
6305 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6306 target A or AAAA records (if not already returned) without resetting the
6307 qualify_single or search_parents options of the DNS resolver. These are
6308 inappropriate in this case because the targets of MX and SRV records must
6309 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6310 target that, when qualified, matched something in the local domain. These
6311 two options are now turned off when doing these lookups.
6313 57. It seems that at least some releases of Reiserfs (which does not have the
6314 concept of a fixed number of inodes) returns zero and not -1 for the
6315 number of available inodes. This interacted badly with check_spool_inodes,
6316 which assumed that -1 was the "no such thing" setting. What I have done is
6317 to check that the total number of inodes is greater than zero before doing
6318 the test of how many are available.
6320 58. When a "warn" ACL statement has a log_message modifier, the message is
6321 remembered, and not repeated. This is to avoid a lot of repetition when a
6322 message has many recipients that cause the same warning to be written.
6323 However, Exim was preserving the list of already written lines for an
6324 entire SMTP session, which doesn't seem right. The memory is now reset if a
6325 new message is started.
6327 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6328 debugging output unless log_rewrite was also set.
6330 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6331 of (int)(handle) when we know that handle contains (void *)(-1).
6333 61. The Exim daemon panic-logs an error return when it closes the incoming
6334 connection. However "connection reset by peer" seems to be common, and
6335 isn't really an error worthy of noting specially, so that particular error
6338 62. When Exim is trying to find all the local interfaces, it used to panic and
6339 die if the ioctl to get the interface flags failed. However, it seems that
6340 on at least one OS (Solaris 9) it is possible to have an interface that is
6341 included in the list of interfaces, but for which you get a failure error
6342 for this call. This happens when the interface is not "plumbed" into a
6343 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6344 failure of the "get flags" call assumes that the interface is down.
6346 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6347 makes life easier for people who are doing arithmetic on fields extracted
6348 from dates, where you often get leading zeros that should not be
6349 interpreted as octal.
6351 64. Added qualify_domain to the redirect router, to override the global
6354 65. If a pathologically long header line contained very many addresses (the
6355 report of this problem mentioned 10 000) and each of them was rewritten,
6356 Exim could use up a very large amount of memory. (It kept on making new
6357 copies of the header line as it rewrote, and never released the old ones.)
6358 At the expense of a bit more processing, the header rewriting function has
6359 been changed so that it no longer eats memory in this way.
6361 66. The generation of the Received: header has been moved from the time that a
6362 message starts to be received, to the time that it finishes. The timestamp
6363 in the Received: header should now be very close to that of the <= log
6364 line. There are two side-effects of this change:
6366 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6367 logged header lines no longer include the local Received: line, because
6368 it has not yet been created. The same applies to a copy of the message
6369 that is returned to a non-SMTP sender when a message is rejected.
6371 (b) When a filter file is tested using -bf, no additional Received: header
6372 is added to the test message. After some thought, I decided that this
6375 This change does not affect the value of $received_for. It is still set
6376 after address rewriting, but before local_scan() is called.
6378 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6380 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6381 gave an unhelpful panic error message, and a defer error. I have managed to
6382 change this behaviour so that it now rejects any supplied certificate,
6383 which seems right, as the list of acceptable certificates is empty.
6385 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6386 gave an unhelpful defer error. I have not managed to make this reject any
6387 supplied certificates, but the error message it gives is "no certificate
6388 supplied", which is not helpful.
6390 70. exigrep's output now also includes lines that are not associated with any
6391 message, but which match the given pattern. Implemented by a patch from
6392 Martin Sluka, which also tidied up the Perl a bit.
6394 71. Recipient callout verification, like sender verification, was using <> in
6395 the MAIL FROM command. This isn't really the right thing, since the actual
6396 sender may affect whether the remote host accepts the recipient or not. I
6397 have changed it to use the actual sender in the callout; this means that
6398 the cache record is now keyed on a recipient/sender pair, not just the
6399 recipient address. There doesn't seem to be a real danger of callout loops,
6400 since a callout by the remote host to check the sender would use <>.
6401 [SEE ABOVE: changed after hitting problems.]
6403 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6404 temporary errors. However, in the case of such a code being given after
6405 the end of a data transmission (i.e. after ".") Exim was failing to write
6406 a retry record for the message. (Yes, there was some broken host that was
6407 actually sending 8xx at this point.)
6409 73. An unknown lookup type in a host list could cause Exim to panic-die when
6410 the list was checked. (An example that provoked this was putting <; in the
6411 middle of a list instead of at the start.) If this happened during a DATA
6412 ACL check, a -D file could be left lying around. This kind of configuration
6413 error no longer causes Exim to die; instead it causes a defer error. The
6414 incident is still logged to the main and panic logs.
6416 74. Buglet left over from Exim 3 conversion. The message "too many messages
6417 in one connection" was written to the rejectlog but not the mainlog, except
6418 when address rewriting (yes!) was being logged.
6420 75. Added write_rejectlog option.
6422 76. When a system filter was run not as root (that is, when system_filter_user
6423 was set), the values of the $n variables were not being returned to the
6424 main process; thus, they were not subsequently available in the $sn
6427 77. Added +return_path_on_delivery log selector.
6429 78. A connection timeout was being treated differently from recipients deferred
6430 when testing hosts_max_try with a message that was older than the host's
6431 retry timeout. (The host should not be counted, thus allowing all hosts to
6432 be tried at least once before bouncing.) This may have been the cause of an
6433 occasionally reported bug whereby a message would remain on the queue
6434 longer than the retry timeout, but would be bounced if a delivery was
6435 forced. I say "may" because I never totally pinned down the problem;
6436 setting up timeout/retry tests is difficult. See also the next item.
6438 79. The ultimate address timeout was not being applied to errors that involved
6439 a combination of host plus message (for example, a timeout on a MAIL
6440 command). When an address resolved to a number of possible hosts, and they
6441 were not all tried for each delivery (e.g. because of hosts_max_try), a
6442 message could remain on the queue longer than the retry timeout.
6444 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6447 81. Fixed an obscure SMTP outgoing bug which required at least the following
6448 conditions: (a) there was another message waiting for the same server;
6449 (b) the server returned 5xx to all RCPT commands in the first message so
6450 that the message was not completed; (c) the server dropped the connection
6451 or gave a negative response to the RSET that Exim sends to abort the
6452 transaction. The observed case was a dropped connection after DATA that had
6453 been sent in pipelining mode. That is, the server had advertised PIPELINING
6454 but was not implementing it correctly. The effect of the bug was incorrect
6455 behaviour, such as trying another host, and this could lead to a crash.
6461 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6462 and daemon.c were passed as pointers to ints; they should have been
6463 pointers to socklen_t variables (which are typically unsigned ints).
6465 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6468 3. Fixed a really odd bug that affected only the testing scheme; patching a
6469 certain fixed string in the binary changed the value of another string that
6470 happened to be identical to the end of the original first string.
6472 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6473 name", it returns that address as the IP address. On some operating
6474 systems (e.g. Solaris), it also passes back the IP address string as the
6475 "host name". However, on others (e.g. Linux), it passes back an empty
6476 string. Exim wasn't checking for this, and was changing the host name to an
6477 empty string, assuming it had been canonicalized.
6479 5. Although rare, it is permitted to have more than one PTR record for a given
6480 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6481 all the names associated with an address, because they do in Solaris.
6482 However, it seems that they do not in Linux for data that comes from the
6483 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6484 I found this out when I moved to a new Linux workstation and tried to run
6485 the Exim test suite.
6487 To get round this problem I have changed the code so that it now does its
6488 own call to the DNS to look up PTR records when searching for a host name.
6489 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6490 addresses that are only in /etc/hosts are still found.
6492 This behaviour is, however, controlled by an option called host_lookup_
6493 order, which defaults to "bydns:byaddr". If people want to use the other
6494 order, or indeed, just use one or the other means of lookup, they can
6495 specify it in this variable.
6497 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6498 some operating systems, this comes back from gethostbyaddr() as an empty
6499 string, and this is what Exim used to test for. However, it seems that in
6500 other systems, "." is yielded. Exim now tests for this case too.
6502 7. The values of check_spool_space and check_log_space are now held internally
6503 as a number of kilobytes instead of an absolute number of bytes. If a
6504 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6505 kilobyte. This means that much larger values can be stored.
6507 8. Exim monitor: an attempt to get the action menu when not actually pointing
6508 at a message produces an empty menu entitled "No message selected". This
6509 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6510 no entries in it ("Shell widget menu has zero width and/or height"). So I
6511 have added a single, blank menu entry in this case.
6513 9. Added ${quote_local_part.
6515 10. MIME decoding is now applied to the contents of Subject: header lines when
6518 11. Now that a reference to $sender_host_address automatically causes a reverse
6519 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6520 host lookup before query-style lookups in lists that might use this
6521 variable. This has therefore been abolished, and the "net-" prefix is no
6522 longer necessary for query-style lookups.
6524 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6525 have been a typo for LFLAGS, so it has been changed.
6527 13. The install script calls Exim with "-C /dev/null" in order to find the
6528 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6529 to be output. However, since Exim outputs its version number before the
6530 error, it didn't break the script. It just looked ugly. I fixed this by
6531 always allowing "-C /dev/null" if the caller is root.
6533 14. Ignore overlarge ACL variable number when reading spool file - insurance
6534 against a later release with more variables having written the file.
6536 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6537 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6538 this, as well as the form without the "IPv6" on the front.
6540 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6541 OS/Makefile-Darwin file.
6543 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6544 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6546 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6547 variable. However, some OS use ints for inodes. I've added cast to long int
6548 to get rid of the compiler warning.
6550 19. I had forgotten to lock out "/../" in configuration file names when
6551 ALT_CONFIG_PREFIX was set.
6553 20. Routers used for verification do not need to specify transports. However,
6554 if such a router generated a host list, and callout was configured, Exim
6555 crashed, because it could not find a port number from the (non-existent)
6556 transport. It now assumes port 25 in this circumstance.
6558 21. Added the -t option to exigrep.
6560 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6561 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6562 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6563 them is set, LOOKUP_LSEARCH is forced.
6565 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6567 24. Added sender and host information to the "rejected by local_scan()" log
6568 line; previously there was no indication of these.
6570 25. Added .include_if_exists.
6572 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6573 Linux. It turns out that not all file systems support this. Apparently some
6574 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6575 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6576 not supported on the file descriptor, is now ignored when Exim is trying to
6577 sync a directory. This applies only to Linux.
6579 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6581 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6582 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6585 29. Added support for maildirsize files from supplied patch (modified a bit).
6587 30. The use of :fail: followed by an empty string could lead Exim to respond to
6588 sender verification failures with (e.g.):
6590 550 Verification failed for <xxx>
6591 550 Sender verify failed
6593 where the first response line was missing the '-' that indicates it is not
6594 the final line of the response.
6596 31. The loop for finding the name of the user that called Exim had a hardwired
6597 limit of 10; it now uses the value of finduser_retries, which is used for
6598 all other user lookups.
6600 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6603 33. Exim was neglecting to zero errno before one call of strtol() when
6604 expanding a string and expecting an integer value. On some systems this
6605 resulted in spurious "integer overflow" errors. Also, it was casting the
6606 result into an int without checking.
6608 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6609 did not work. The code looks as if it has *never* worked, though it appears
6610 to have been documented since at least release 1.62. I have made it work.
6612 35. The "timeout_DNS" error in retry rules, also documented since at least
6613 1.62, also never worked. As it isn't clear exactly what this means, and
6614 clearly it isn't a major issue, I have abolished the feature by treating it
6615 as "timeout", and writing a warning to the main and panic logs.
6617 36. The display of retry rules for -brt wasn't always showing the error code
6620 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6621 timeout_connect_A, timeout_connect_MX.
6623 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6624 to the empty sender.
6626 39. The daemon was not analysing the content of -oX till after it had closed
6627 stderr and disconnected from the controlling terminal. This meant that any
6628 syntax errors were only noted on the panic log, and the return code from
6629 the command was 0. By re-arranging the code a little, I've made the
6630 decoding happen first, so such errors now appear on stderr, and the return
6631 code is 1. However, the actual setting up of the sockets still happens in
6632 the disconnected process, so errors there are still only recorded on the
6635 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6636 connections (as happens on some IP stacks) was logged at start up time as
6637 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6638 it from "IPv6 and IPv4", which means that two separate sockets are being
6641 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6642 says whether AF_INET or AF_INET6 was passed as an argument.
6644 42. Exiwhat output was messed up when time zones were included in log
6647 43. Exiwhat now gives more information about the daemon's listening ports,
6648 and whether -tls-on-connect was used.
6650 44. The "port" option of the smtp transport is now expanded.
6652 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6653 silently ignored. Now an error message is written to the main and panic
6656 46. There's a new ACL modifier called "logwrite" which writes to a log file
6657 as soon as it is encountered.
6659 47. Added $local_user_uid and $local_user_gid at routing time.
6661 48. Exim crashed when trying to verify a sender address that was being
6664 49. Exim was recognizing only a space character after ".include". It now also
6665 recognizes a tab character.
6667 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6668 extracting the relevant information from the specification. The man page no
6669 longer contains scrambled data for the -d option, and I've added a section
6670 at the front about calling Exim under different names.
6672 51. Added "extra_headers" argument to the "mail" command in filter files.
6674 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6677 53. Installed eximstats 1.29.
6679 54. Added transport_filter_timeout as a generic transport option.
6681 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6682 Cc: header lines. This was required by RFC 822, but it not required by RFC
6685 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6686 incoming messages that did not have them. Now it does so only if the
6687 message originates locally, that is, if there is no associated remote host
6688 address. When Resent- header lines are present, this applies to the Resent-
6689 lines rather than the non-Resent- lines.
6691 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6692 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6694 58. Messages for configuration errors now include the name of the main
6695 configuration file - useful now that there may be more than one file in a
6696 list (.included file names were always shown).
6698 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6699 for those rare installations that do not start the daemon as root or run it
6700 setuid root. I've cut out the call to initgroups() if the daemon is not
6703 60. The Exim user and group can now be bound into the binary as text strings
6704 that are looked up at the start of Exim's processing.
6706 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6708 62. Added $mailstore_basename variable.
6710 63. Installed patch to sieve.c from Michael Haardt.
6712 64. When Exim failed to open the panic log after failing to open the main log,
6713 the original message it was trying to log was written to stderr and debug
6714 output, but if they were not available (the usual case in production), it
6715 was lost. Now it is written to syslog before the two lines that record the
6716 failures to open the logs.
6718 65. Users' Exim filters run in subprocesses under the user's uid. It is
6719 possible for a "deliver" command or an alias in a "personal" command to
6720 provoke an address rewrite. If logging of address rewriting is configured,
6721 this fails because the process is not running as root or exim. There may be
6722 a better way of dealing with this, but for the moment (because 4.30 needs
6723 to be released), I have disabled address rewrite logging when running a
6724 filter in a non-root, non-exim process.
6730 1. The buildconfig auxiliary program wasn't quoting the value set for
6731 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6732 not defined. This bug was masked in 4.22 by the effect that was fixed in
6735 2. Some messages that were rejected after a message id was allocated were
6736 shown as "incomplete" by exigrep. It no longer does this for messages that
6737 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6739 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6740 have allow_domain_literals set, the ID did not get logged in the <= line.
6741 Domain literals are now always recognized in Message-ID: header lines.
6743 4. The first argument for a ${extract expansion item is the key name or field
6744 number. Leading and trailing spaces in this item were not being ignored,
6745 causing some misleading effects.
6747 5. When deliver_drop_privilege was set, single queue runner processes started
6748 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6749 same command in the process it spins off) were not dropping privilege.
6751 6. When the daemon running as "exim" started a queue runner, it always
6752 re-executed Exim in the spun-off process. This is a waste of effort when
6753 deliver_drop_privilege is set. The new process now just calls the
6754 queue-runner function directly.
6760 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6763 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6764 ignored. Though the use of -oP was forcing the writing of a pid file, it
6765 was always written to the default place.
6767 3. If the message "no IP address found for host xxxx" is generated during
6768 incoming verification, it is now followed by identification of the incoming
6769 connection (so you can more easily find what provoked it).
6771 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6773 5. Added some features to "harden" Exim a bit more against certain attacks:
6775 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6776 be put in Local/Makefile. This is like the never_users runtime option,
6777 but it cannot be overridden. The default setting is "root".
6779 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6780 prefix string with which any file named in a -C command line option
6783 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6784 is retained for -C and -D only if the caller of Exim is root. Without
6785 it, the exim user may also use -C and -D and retain privilege.
6787 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6788 command line option is disabled.
6790 6. Macro names set by the -D option must start with an upper case letter, just
6791 like macro names defined in the configuration file.
6793 7. Added "dereference=" facility to LDAP.
6795 8. Two instances of the typo "uknown" in the source files are fixed.
6797 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6798 the Configure-Makefile script screwed up while processing it.
6800 10. Incorporated PCRE 4.4.
6802 11. The SMTP synchronization check was not operating right at the start of an
6803 SMTP session. For example, it could not catch a HELO sent before the client
6804 waited for the greeting. There is now a check for outstanding input at the
6805 point when the greeting is written. Because of the duplex, asynchronous
6806 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6807 way, but not yet received, when the check is performed.
6809 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6810 on TCP/IP sockets, because this apparently causes some broken clients to
6813 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6814 unchanged) from the Cygwin maintainer.
6816 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6817 was supported instead of testing for mbx. Effectively a typo.
6819 15. The spa authenticator server code was not checking that the input it
6820 received was valid base64.
6822 16. The debug output line for the "set" modifier in ACLs was not showing the
6823 name of the variable that was being set.
6825 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6827 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6828 lookup on its own. Something else was needed to trigger the lookup. For
6829 example, a match in host_lookup or the need for a host name in a host list.
6830 Now, if $sender_host_name is referenced and the host name has not yet been
6831 looked up, a lookup is performed. If the lookup fails, the variable remains
6832 empty, and $host_lookup_failed is set to "1".
6834 19. Added "eqi" as a case-independent comparison operator.
6836 20. The saslauthd authentication condition could segfault if neither service
6837 nor realm was specified.
6839 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6840 error message that was logged was misleading, and incoming SMTP
6841 connections were dropped. The message is now more accurate, and temporary
6842 errors are given to SMTP connections.
6844 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6845 (except RSET) with a 421 error, until QUIT is received. However, it was
6846 failing to send a response to QUIT.
6848 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6849 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6850 ACL is now run after the helo_try_verify_hosts code.
6852 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6853 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6854 case-independent, so other case variants are also recognized). Apparently
6855 some systems use these upper case variants.
6857 25. If more than two messages were waiting for the same host, and a transport
6858 filter was specified for the transport, Exim sent two messages over the
6859 same TCP/IP connection, and then failed with "socket operation on non-
6860 socket" when it tried to send the third.
6862 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6864 27. The extern definition of crypt16() in expand.c was not being excluded when
6865 the OS had its own crypt16() function.
6867 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6868 as a preferred synonym for return_size_limit, both as an option and as an
6871 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6873 30. Changed the default configuration ACL to relax the local part checking rule
6874 for addresses that are not in any local domains. For these addresses,
6875 slashes and pipe symbols are allowed within local parts, but the sequence
6876 /../ is explicitly forbidden.
6878 31. SPA server authentication was not clearing the challenge buffer before
6881 32. log_message in a "warn" ACL statement was writing to the reject log as
6882 well as to the main log, which contradicts the documentation and doesn't
6883 seem right (because no rejection is happening). So I have stopped it.
6885 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6886 However, I am unable to do any testing of this.
6888 34. Fixed an infelicity in the appendfile transport. When checking directories
6889 for a mailbox, to see if any needed to be created, it was accidentally
6890 using path names with one or more superfluous leading slashes; tracing
6891 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6893 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6894 contents are added to the log line that is written for every discarded
6895 recipient. (Previously a log_message setting was ignored.)
6897 36. The ${quote: operator now quotes the string if it is empty.
6899 37. The install script runs exim in order to find its version number. If for
6900 some reason other than non-existence or emptiness, which it checks, it
6901 could not run './exim', it was installing it with an empty version number,
6902 i.e. as "exim-". This error state is now caught, and the installation is
6905 38. An argument was missing from the function that creates an error message
6906 when Exim fails to connect to the socket for saslauthd authentication.
6907 This could cause Exim to crash, or give a corrupted message.
6909 39. Added isip, isip4, and isip6 to ${if conditions.
6911 40. The ACL variables $acl_xx are now saved with the message, and can be
6912 accessed later in routers, transports, and filters.
6914 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6915 strings in the file are not string-expanded.
6917 42. If a MAIL command specified a SIZE value that was too large to fit into an
6918 int variable, the check against message_size_limit failed. Such values are
6919 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6920 day this will have to be increased, but I don't think I want to be around
6921 when emails are that large.
6928 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6929 iconv() is not standard in FreeBSD.
6931 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6932 IPv6 enabled. The observed symptom was a segmentation fault on return from
6933 the function os_common_find_running_interfaces() in src/os.c.
6935 3. In the check_special_case() function in daemon.c I had used "errno" as an
6936 argument name, which causes warnings on some systems. This was basically a
6937 typo, since it was named "eno" in the comments!
6939 4. The code that waits for the clock to tick (at a resolution of some fraction
6940 of a second) so as to ensure message-id uniqueness was always waiting for
6941 at least one whole tick, when it could have waited for less. [This is
6942 almost certainly not relevant at current processor speeds, where it is
6943 unlikely to ever wait at all. But we try to future-proof.]
6945 5. The function that sleeps for a time interval that includes fractions of a
6946 second contained a race. It did not block SIGALRM between setting the
6947 timer, and suspending (a couple of lines later). If the interval was short
6948 and the sigsuspend() was delayed until after it had expired, the suspension
6949 never ended. On busy systems this could lead to processes getting stuck for
6952 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6953 process, before it forks any delivery processes. The open lookup caching
6954 mechanism meant that the open file or database connection was passed into
6955 the delivery process. The problem was that delivery processes always tidy
6956 up cached lookup data. This could cause a problem for the next delivery
6957 process started by the queue runner, because the external queue runner
6958 process does not know about the closure. So the next delivery process
6959 still has data in the lookup cache. In the case of a file lookup, there was
6960 no problem because closing a file descriptor in a subprocess doesn't affect
6961 the parent. However, if the lookup was caching a connection to a database,
6962 the connection was closed, and the second delivery process was likely to
6963 see errors such as "PGSQL: query failed: server closed the connection
6964 unexpectedly". The problem has been fixed by closing all cached lookups
6965 in a queue runner before running a delivery process.
6967 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6968 seem to have the "const" qualifier which it has on other OS. I've
6971 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6972 *of the same type* (client or server) with the same public name that an
6973 error should be diagnosed.
6975 9. When Exim looked up a host name for an IP address, but failed to find the
6976 original IP address when looking up the host name (a safety check), it
6977 output the message "<ip address> does not match any IP for NULL", which was
6978 confusing, to say the least. The bug was that the host name should have
6979 appeared instead of "NULL".
6981 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6982 user that is built into the binary, and the -C or -D options is used, root
6983 privilege is dropped before the configuration file is read. In addition,
6984 logging is switched to stderr instead of the normal log files. If the
6985 configuration then re-defines the Exim user, the unprivileged environment
6986 is probably not what is expected, so Exim logs a panic warning message (but
6989 However, if deliver_drop_privilege is set, the unprivileged state may well
6990 be exactly what is intended, so the warning has been cut out in that case,
6991 and Exim is allowed to try to write to its normal log files.
6997 1. smtp_return_error_details was not giving details for temporary sender
6998 or receiver verification errors.
7000 2. Diagnose a configuration error if two authenticators have the same public
7003 3. Exim used not to create the message log file for a message until the first
7004 delivery attempt. This could be confusing when incoming messages were held
7005 for policy or load reasons. The message log file is now created at the time
7006 the message is received, and an initial "Received" line is written to it.
7008 4. The automatically generated man page for command line options had a minor
7009 bug that caused no ill effects; however, a more serious problem was that
7010 the procedure for building the man page automatically didn't always
7011 operate. Consequently, release 4.20 contains an out-of-date version. This
7012 shouldn't happen again.
7014 5. When building Exim with embedded Perl support, the script that builds the
7015 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7016 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7018 6. The freeze_tell option was not being used for messages that were frozen on
7019 arrival, either by an ACL or by local_scan().
7021 7. Added the smtp_incomplete_transaction log selector.
7023 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7024 was accepting AUTH without a new EHLO.
7026 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7027 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7030 10. Logging of TCP/IP connections (when configured) now happens in the main
7031 daemon process instead of the child process, so that the TCP/IP connection
7032 count is more accurate (but it can never be perfect).
7034 11. The use of "drop" in a nested ACL was not being handled correctly in the
7035 outer ACL. Now, if condition failure induced by the nested "drop" causes
7036 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7037 or "require"), the connection is dropped.
7039 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7040 that yield "discard" can now be used with an "accept" or a "discard" verb,
7041 but an error is generated for any others (because I can't see a useful way
7042 to define what should happen).
7044 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7045 are now processed in the same way as lines in the Exim configuration file.
7046 In particular, continuation lines are supported.
7048 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7050 15. Added -ti meaning -t -i.
7052 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7053 domains, and warn by logging if others are found.
7055 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7056 returned from GIFCONF when Exim is trying to find the list of interfaces on
7057 a host. The code in os.c has been modified to copy each ifreq to an aligned
7058 structure in all cases.
7060 Also, in some cases, the returned ifreq's were being copied to a 'struct
7061 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7062 means the last couple of bytes of an IPv6 address could be chopped if the
7063 ifreq contained only a normal sockaddr (14 bytes storage).
7065 18. Named domain lists were not supported in the hosts_treat_as_local option.
7066 An entry such as +xxxx was not recognized, and was treated as a literal
7069 19. Ensure that header lines added by a DATA ACL are included in the reject log
7070 if the ACL subsequently rejects the message.
7072 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7073 MD5 (which is deprecated).
7075 21. When testing a filter file using -bf, Exim was writing a message when it
7076 took the sender from a "From " line in the message, but it was not doing so
7077 when it took $return_path from a Return-Path: header line. It now does.
7079 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7080 with a valid header line field name (a series of printing characters
7081 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7083 23. Changed "disc" in the source to "disk" to conform to the documentation and
7084 the book and for uniformity.
7086 24. Ignore Sendmail's -Ooption=value command line item.
7088 25. When execve() failed while trying to run a command in a pipe transport,
7089 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7090 could be confused with a return value of 69 from the command itself. This
7091 has been changed to 127, the value the shell returns if it is asked to run
7092 a non-existent command. The wording for the related log line suggests a
7093 non-existent command as the problem.
7095 26. If received_header_text expands to an empty string, do not add a Received:
7096 header line to the message. (Well, it adds a token one on the spool, but
7097 marks it "old" so that it doesn't get used or transmitted.)
7099 27. Installed eximstats 1.28 (addition of -nt option).
7101 28. There was no check for failure on the call to getsockname() in the daemon
7102 code. This can fail if there is a shortage of resources on the system, with
7103 ENOMEM, for example. A temporary error is now given on failure.
7105 29. Contrary to the C standard, it seems that in some environments, the
7106 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7107 program. Exim now does this explicitly; it affects the formatting of
7108 timestamps using strftime().
7110 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7111 complaints. I've now initialized all the variables, to avoid this.
7113 32. Header lines added by a system filter were not being "seen" during
7114 transport-time rewrites.
7116 33. The info_callback() function passed to OpenSSL is set up with type void
7117 (*)(SSL *, int, int), as described somewhere. However, when calling the
7118 function (actually a macro) that sets it up, the type void(*)() is
7119 expected. I've put in a cast to prevent warnings from picky compilers.
7121 34. If a DNS black list lookup found a CNAME record, but there were no A
7122 records associated with the domain it pointed at, Exim crashed.
7124 35. If a DNS black list lookup returned more than one A record, Exim ignored
7125 all but the first. It now scans all returned addresses if a particular IP
7126 value is being sought. In this situation, the contents of the
7127 $dnslist_value variable are a list of all the addresses, separated by a
7130 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7131 to accept a host name and all its aliases if the forward lookup for any of
7132 them yielded the IP address of the incoming connection. Now it accepts only
7133 those names whose forward lookup yields the correct IP address. Any other
7134 names are discarded. This closes a loophole whereby a rogue DNS
7135 administrator could create reverse DNS records to break through a
7136 wildcarded host restriction in an ACL.
7138 37. If a user filter or a system filter that ran in a subprocess used any of
7139 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7140 the wrong values were passed to the pipe command ($thisaddress had the
7141 value of $0, $0 had the value of $1, etc). This bug was introduced by
7142 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7144 38. Improved the line breaking for long SMTP error messages from ACLs.
7145 Previously, if there was no break point between 40 and 75 characters, Exim
7146 left the rest of the message alone. Two changes have been made: (a) I've
7147 reduced the minimum length to 35 characters; (b) if it can't find a break
7148 point between 35 and 75 characters, it looks ahead and uses the first one
7149 that it finds. This may give the occasional overlong line, but at least the
7150 remaining text gets split now.
7152 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7153 file descriptors might be low, and that setting 1000 would always raise it.
7154 It turns out that in some environments, the limit is already over 1000 and
7155 that lowering it causes trouble. So now Exim takes care not to decrease it.
7157 40. When delivering a message, the value of $return_path is set to $sender_
7158 address at the start of routing (routers may change the value). By an
7159 oversight, this default was not being set up when an address was tested by
7160 -bt or -bv, which affected the outcome if any router or filter referred to
7163 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7164 log only when "message" or "log_message" are set. However, if one of the
7165 conditions was an address verification, or a call to a nested ACL, the
7166 messages generated by the underlying test were being passed through. This
7167 no longer happens. The underlying message is available in $acl_verify_
7168 message for both "message" and "log_message" expansions, so it can be
7169 passed through if needed.
7171 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7172 new expansion $bh_ to give the encoded byte string without charset
7173 translation. Translation happens only if iconv() is available; HAVE_ICONV
7174 indicates this at build time. HEADERS_CHARSET gives the charset to
7175 translate to; headers_charset can change it in the configuration, and
7176 "headers charset" can change it in an individual filter file.
7178 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7179 that creates RFC 2047 encoded "words" labels them as that charset instead
7180 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7181 expansion operator; (ii) when Exim creates a From: line for a local
7182 message; (iii) when a header line is rewritten to include a "phrase" part.
7184 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7185 buggy, causing it to skip the first lines of messages whose message ID
7186 ended in 'D'. This would not have bitten before Exim release 4.14, because
7187 message IDs were unlikely to end in 'D' before then. The effect was to have
7188 incorrect size information for certain domains.
7190 45. #include "config.h" was missing at the start of the crypt16.c module. This
7191 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7194 46. If there was a timeout during a "random" callout check, Exim treated it as
7195 a failure of the random address, and carried on sending RSET and the real
7196 address. If the delay was just some slowness somewhere, the response to the
7197 original RCPT would be taken as a response to RSET and so on, causing
7198 mayhem of various kinds.
7200 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7201 when I implemented it. It didn't allow for the fact that some option values
7202 may legitimately be negative (e.g. size_addition), and it didn't even do
7203 the right test for positive values.
7205 48. Domain names in DNS records are case-independent. Exim always looks them up
7206 in lower case. Some resolvers return domain names in exactly the case they
7207 appear in the zone file, that is, they may contain uppercase letters. Not
7208 all resolvers do this - some return always lower case. Exim was treating a
7209 change of case by a resolver as a change of domain, similar to a widening
7210 of a domain abbreviation. This triggered its re-routing code and so it was
7211 trying to route what was effectively the same domain again. This normally
7212 caused routing to fail (because the router wouldn't handle the domain
7213 twice). Now Exim checks for this case specially, and just changes the
7214 casing of the domain that it ultimately uses when it transmits the message
7217 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7220 50. If a filter generated a file delivery with a non-absolute name (possible if
7221 no home directory exists for the router), the forbid_file option was not
7224 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7225 the existing equality matching.
7227 52. Exim was using ints instead of ino_t variables in some places where it was
7228 dealing with inode numbers.
7230 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7231 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7232 called TMPDIR, and if it finds it is different, it changes its value.
7234 54. The smtp_printf() function is now made available to local_scan() so
7235 additional output lines can be written before returning. There is also an
7236 smtp_fflush() function to enable the detection of a dropped connection.
7237 The variables smtp_input and smtp_batched_input are exported to
7240 55. Changed the default runtime configuration: the message "Unknown user"
7241 has been removed from the ACL, and instead placed on the localuser router,
7242 using the cannot_route_message feature. This means that any verification
7243 failures that generate their own messages won't get overridden. Similarly,
7244 the "Unrouteable address" message that was in the ACL for unverifiable
7245 relay addresses has also been removed.
7247 56. Added hosts_avoid_esmtp to the smtp transport.
7249 57. The exicyclog script was not checking for the esoteric option
7250 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7251 will work only if exicyclog is run under the appropriate euid.
7253 58. Following a discussion on the list, the rules by which Exim recognises line
7254 endings on incoming messages have been changed. The -dropcr and drop_cr
7255 options are now no-ops, retained only for backwards compatibility. The
7256 following line terminators are recognized: LF CRLF CR. However, special
7257 processing applies to CR:
7259 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7260 nor a local message in the state where . is a terminator.
7262 (ii) If a bare CR is encountered in a header line, an extra space is added
7263 after the line terminator so as not to end the header. The reasoning
7264 behind this is that bare CRs in header lines are most likely either
7265 to be mistakes, or people trying to play silly games.
7267 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7268 was being incorrectly given as 18 bytes larger than it should have been.
7269 This is a VOB (very old bug).
7271 60. This may never have affected anything current, but just in case it has:
7272 When the local host is found other than at the start of a list of hosts,
7273 the local host, those with the same MX, and any that follow, are discarded.
7274 When the list in question was part of a longer list of hosts, the following
7275 hosts (not currently being processed) were also being discarded. This no
7276 longer happens. I'm not sure if this situation could ever has previously
7279 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7282 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7283 header. This is something that is recommended in a new Internet Draft, and
7284 is something that is documented as being done by Sendmail. There are two
7285 possible values. For messages generated by the autoreply transport, Exim
7288 Auto-Submitted: auto-replied
7290 whereas for all other generated messages (e.g. bounces) it adds
7292 Auto-Submitted: auto-generated
7294 63. The "personal" condition in filters now includes a test for the
7295 Auto-Submitted: header. If it contains the string "auto-" the message it
7296 not considered personal.
7298 64. Added rcpt_include_affixes as a generic transport option.
7300 65. Added queue_only_override (default true).
7302 66. Added the syslog_duplication option.
7304 67. If what should have been the first header line of a message consisted of
7305 a space followed by a colon, Exim was mis-interpreting it as a header line.
7306 It isn't of course - it is syntactically invalid and should therefore be
7307 treated as the start of the message body. The misbehaviour could have
7308 caused a number of strange effects, including loss of data in subsequent
7309 header lines, and spool format errors.
7311 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7312 client host had authenticated. This control can now be exercised by an ACL
7313 for more flexibility.
7315 69. By default, callouts do not happen when testing with -bh. There is now a
7316 variant, -bhc, which does actually run the callout code, including
7317 consulting and updating the callout cache.
7319 70. Added support for saslauthd authentication, courtesy of Alexander
7322 71. If statvfs() failed on the spool or log directories while checking their
7323 size for availability, Exim confusingly gave the error "space shortage".
7324 Furthermore, in debugging mode it crashed with a floating point exception.
7325 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7326 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7327 serious problem, Exim now writes to the main and panic logs when this
7328 happens, with details of the failure. It then refuses to accept the
7329 incoming message, giving the message "spool directory problem" or "log
7330 directory problem" with a 421 code for SMTP messages.
7332 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7333 0, 1, and 2 exist, because some OS complain for execs without them (see
7334 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7335 descriptors. However, the code omitted to check that the open succeeded,
7336 causing mysterious errors if for some reason the permissions on /dev/null
7337 got screwed. Now Exim writes a message to the main and panic logs, and
7338 bombs out if it can't open /dev/null.
7340 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7341 interact so that it is all more flexible. It is supposed to remain
7342 backwards compatible. Also added extra_local_interfaces.
7344 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7345 to bomb out with an assertion failure - to the client this appears as a
7346 connection drop. This problem occurs in the part of the code that was taken
7347 from the Samba project. Fortunately, the assertion is in a very simple
7348 function, so I have fixed this by reproducing the function inline in the
7349 one place where it is called, and arranging for authentication to fail
7350 instead of killing the process with assert().
7352 75. The SPA client code was not working when the server requested OEM rather
7353 than Unicode encoding.
7355 76. Added code to make require_files with a specific uid setting more usable in
7356 the case where statting the file as root fails - usually a non-root-mounted
7357 NFS file system. When this happens and the failure is EACCES, Exim now
7358 forks a subprocess and does the per-uid checking as the relevant uid.
7360 77. Added process_log_path.
7362 78. If log_file_path was not explicitly set, a setting of check_log_space or
7363 check_log_inodes was ignored.
7365 79. If a space check for the spool or log partitions fails, the incident is now
7366 logged. Of course, in the latter case the data may get lost...
7368 80. Added the %p formatting code to string_format() so that it can be used to
7369 print addresses in debug_print(). Adjusted all the address printing in the
7370 debugging in store.c to use %p rather than %d.
7372 81. There was a concern that a line of code in smtp_in.c could overflow a
7373 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7374 initially expressed, the concern was not well-founded, because trailing
7375 spaces are removed early. However, if the trailing spaces were followed by
7376 a NULL, they did not get removed, so the overflow was possible. Two fixes
7379 (a) I re-wrote the offending code in a cleaner fashion.
7380 (b) If an incoming SMTP command contains a NULL character, it is rejected
7383 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7384 runs initgroups(), so that if the Exim user is in any additional groups,
7385 they will be used during message reception.
7391 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob