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/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
17 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
18 buffer overrun for (non-chunking) other transports.
24 JH/01 Remove code calling the customisable local_scan function, unless a new
25 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
27 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
28 non-signal-safe functions being used.
30 JH/03 Bug 2269: When presented with a received message having a stupidly large
31 number of DKIM-Signature headers, disable DKIM verification to avoid
32 a resource-consumption attack. The limit is set at twenty.
34 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
35 report of oldest_pass in ${authres } in consequence, and separate out
36 some descriptions of reasons for verification fail.
38 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
39 files in the spool were present and unlocked. A queue-runner could spot
40 them, resulting in a duplicate delivery. Fix that by doing the unlock
41 after the unlink. Investigation by Tim Stewart. Take the opportunity to
42 add more error-checking on spoolfile handling while that code is being
45 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
46 No known attacks, no CVE, this is defensive hardening.
48 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
49 a queue-runner could start a delivery while other operations were ongoing.
50 Cutthrough delivery was a common victim, resulting in duplicate delivery.
51 Found and investigated by Tim Stewart. Fix by using the open message data
52 file handle rather than opening another, and not locally closing it (which
53 releases a lock) for that case, while creating the temporary .eml format
54 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
56 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
57 $sender_verify_failure/$recipient_verify_failure to "random".
59 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
62 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
63 Previously this would segfault.
65 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
68 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
69 like zero, since the resolver should be doing this for us, But we need one
70 as a CNAME but no MX presence gets the CNAME returned; we need to check
71 that doesn't point to an MX to declare it "no MX returned" rather than
72 "error, loop". A new main option is added so the older capability of
73 following some limited number of chain links is maintained.
75 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
77 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
78 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
79 now-deprecated earlier definitions used only specified the range up to TLS
80 1.2 (in the older-version library docs).
82 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
84 JH/15 Rework TLS client-side context management. Stop using a global, and
85 explicitly pass a context around. This enables future use of TLS for
86 connections to service-daemons (eg. malware scanning) while a client smtp
87 connection is using TLS; with cutthrough connections this is quite likely.
89 JH/16 Fix ARC verification to do AS checks in reverse order.
91 JH/17 Support a "tls" option on the ${readsocket } expansion item.
93 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
94 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
95 Previously the "utf8" would be re-prepended for every additional message.
97 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
98 Previously thery were accepted, resulting in issues when attempting to
99 forward messages to a non-supporting MTA.
101 PP/02 Let -n work with printing macros too, not just options.
103 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
104 one parent address was copied, and bogus data was used at delivery-logging
105 time. Either a crash (after delivery) or bogus log data could result.
106 Discovery and analysis by Tim Stewart.
108 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
109 Previously if the string ended mid-character, we did not insert the
110 promised '?' replacement.
112 PP/04 Documentation: current string operators work on bytes, not codepoints.
114 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
115 should pack well giving a smaller memory footprint so better caching and
116 therefore performance. Group the declarations where this can't be done so
117 that the byte-sized flag variables are not interspersed among pointer
118 variables, giving a better chance of good packing by the compiler.
120 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
121 non-null, to avoid issues with sites running BATV. Previously reports were
122 sent with an empty envelope sender so looked like bounces.
124 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
125 The ignore_error flag wasn't being returned from the filter subprocess so
126 was not set for later routers. Investigation and fix by Matthias Kurz.
128 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
129 and a msg:complete for the whole, when a message is manually removed using
130 -Mrm. Developement by Matthias Kurz, hacked on by JH.
132 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
133 a "Gnu special" function, asprintf() in the DB utility binary builds; I
134 hope that is portable enough.
136 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
137 requiring a known-CA anchor certificate; make it now rely entirely on the
138 TLSA as an anchor. Checking the name on the leaf cert against the name
139 on the A-record for the host is still done for TA (but not for EE mode).
141 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
142 included in delivery lines for non-proxied connections, when compiled with
143 SUPPORT_SOCKS and running with proxy logging enabled.
145 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
146 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
147 move the existing event to fire before the normal logging of message
148 failure so that custom logging is bracketed by normal logging.
150 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
151 msg:fail:internal event. Developement by Matthias Kurz.
153 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
154 far too small for todays use of crypto signatures stored there. Go all
155 the way to the max DNS message size of 64kB, even though this might be
156 overmuch for IOT constrained device use.
158 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
159 copy a string over itself. The library routine is documented as not
160 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
162 JH/32 For main options check_spool_space and check_inode_space, where the
163 platform supports 64b integers, support more than the previous 2^31 kB
164 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
165 the previous G, M, k.
167 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
168 $authenticated_fail_id variable on authentication failure. Previously
171 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
172 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
173 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
174 GNUTLS_SEC_PARAM_MEDIUM.
176 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
177 side. Previously we would continue as if no SNI had been received.
179 JH/36 Harden the handling of string-lists. When a list consisted of a sole
180 "<" character, which should be a list-separator specification, we walked
181 off past the nul-terimation.
183 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
184 causes) even when the retry time is not yet met. Previously they were
185 not, meaning that when (say) an account was over-quota and temp-rejecting,
186 and multiple senders' messages were queued, only one sender would get
187 notified on each configured delay_warning cycle.
189 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
191 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
192 cipher-suites, an error can be left on the stack even for a succeeding
193 accept; this results in impossible error messages when a later operation
196 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
197 return error codes indicating retry. Under TLS1.3 this becomes required.
199 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
200 it only wrote the new authenticators, resulting in a lack of tracking of
201 peer changes of ESMTP extensions until the next cache flush.
203 JH/41 Fix the loop reading a message header line to check for integer overflow,
204 and more-often against header_maxsize. Previously a crafted message could
205 induce a crash of the recive process; now the message is cleanly rejected.
207 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
208 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
214 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
215 When redis_servers is set to a list of > 1 element, and the Redis servers
216 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
217 case of MOVED into a DEFER case instead, thus moving the query onto the
218 next server in the list. For a cluster of N elements, all N servers must
219 be defined in redis_servers.
221 GF/02 Catch and remove uninitialized value warning in exiqsumm
222 Check for existence of @ARGV before looking at $ARGV[0]
224 JH/01 Replace the store_release() internal interface with store_newblock(),
225 which internalises the check required to safely use the old one, plus
226 the allocate and data copy operations duplicated in both (!) of the
227 extant use locations.
229 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
230 modifier. This matches the restriction on the commandline.
232 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
233 Previously only the last row was returned.
235 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
236 we assumed that tags in the header were well-formed, and parsed the
237 element content after inspecting only the first char of the tag.
238 Assumptions at that stage could crash the receive process on malformed
241 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
242 While running the DKIM ACL we operate on the Permanent memory pool so that
243 variables created with "set" persist to the DATA ACL. Also (at any time)
244 DNS lookups that fail create cache records using the Permanent pool. But
245 expansions release any allocations made on the current pool - so a dnsdb
246 lookup expansion done in the DKIM ACL releases the memory used for the
247 DNS negative-cache, and bad things result. Solution is to switch to the
248 Main pool for expansions.
249 While we're in that code, add checks on the DNS cache during store_reset,
250 active in the testsuite.
251 Problem spotted, and debugging aided, by Wolfgang Breyha.
253 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
254 When none of the hosts presented to a transport match an already-open
255 connection, close it and proceed with the list. Previously we would
256 queue the message. Spotted by Lena with Yahoo, probably involving
259 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
260 Previously a spurious "250 OK id=" response was appended to the proper
263 JH/08 The "support for" informational output now, which built with Content
264 Scanning support, has a line for the malware scanner interfaces compiled
265 in. Interface can be individually included or not at build time.
267 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
268 by the template makefile "src/EDITME". The "STREAM" support for an older
269 ClamAV interface method is removed.
271 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
272 rows affected is given instead).
274 JH/11 The runtime Berkeley DB library version is now additionally output by
275 "exim -d -bV". Previously only the compile-time version was shown.
277 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
278 SMTP connection. Previously, when one had more receipients than the
279 first, an abortive onward connection was made. Move to full support for
280 multiple onward connections in sequence, handling cutthrough connection
281 for all multi-message initiating connections.
283 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
284 routers. Previously, a multi-recipient message would fail to match the
285 onward-connection opened for the first recipient, and cause its closure.
287 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
288 a timeout on read on a GnuTLS initiating connection, resulting in the
289 initiating connection being dropped. This mattered most when the callout
290 was marked defer_ok. Fix to keep the two timeout-detection methods
293 JH/15 Relax results from ACL control request to enable cutthrough, in
294 unsupported situations, from error to silently (except under debug)
295 ignoring. This covers use with PRDR, frozen messages, queue-only and
298 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
300 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
301 metadata, resulting in a crash in free().
303 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
304 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
305 Broken also in d185889f4, with init system revamp.
307 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
308 Previously we abruptly closed the connection after reading a malware-
309 found indication; now we go on to read the "scan ok" response line,
312 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
313 ACL. Previously, a crash would result.
315 JH/19 Speed up macro lookups during configuration file read, by skipping non-
316 macro text after a replacement (previously it was only once per line) and
317 by skipping builtin macros when searching for an uppercase lead character.
319 JH/20 DANE support moved from Experimental to mainline. The Makefile control
320 for the build is renamed.
322 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
323 was allocated for every new TLS startup, meaning one per message. Fix
324 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
326 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
327 reported the original. Fix to report (as far as possible) the ACL
328 result replacing the original.
330 JH/23 Fix memory leak during multi-message connections using STARTTLS under
331 OpenSSL. Certificate information is loaded for every new TLS startup,
332 and the resources needed to be freed.
334 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
336 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
339 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
340 DATA response info to the (existing) per-recipient response info for
341 the "C=" log element. It can have useful tracking info from the
342 destination system. Patch from Simon Arlott.
344 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
345 length value. Previously this would segfault.
347 HS/02 Support Avast multiline protoocol, this allows passing flags to
348 newer versions of the scanner.
350 JH/28 Ensure that variables possibly set during message acceptance are marked
351 dead before release of memory in the daemon loop. This stops complaints
352 about them when the debug_store option is enabled. Discovered specifically
353 for sender_rate_period, but applies to a whole set of variables.
354 Do the same for the queue-runner and queue-list loops, for variables set
355 from spool message files. Do the same for the SMTP per-message loop, for
356 certain variables indirectly set in ACL operations.
358 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
359 as a multi-recipient message from a mailinglist manager). The coding had
360 an arbitrary cutoff number of characters while checking for more input;
361 enforced by writing a NUL into the buffer. This corrupted long / fast
362 input. The problem was exposed more widely when more pipelineing of SMTP
363 responses was introduced, and one Exim system was feeding another.
364 The symptom is log complaints of SMTP syntax error (NUL chars) on the
365 receiving system, and refused recipients seen by the sending system
366 (propating to people being dropped from mailing lists).
367 Discovered and pinpointed by David Carter.
369 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
370 replaced by the ${authresults } expansion.
372 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
374 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
375 allows proper process termination in container environments.
377 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
378 Previously the "final dot" had a newline after it; ensure it is CR,LF.
380 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
381 and "err_perm", deprecated since 4.83 when the RFC-defined words
382 "temperror" and "permerror" were introduced.
384 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
385 transport-filters or DKIM-signing. The restriction was lost in the
386 consolidation of verify-callout and delivery SMTP handling.
387 Extend the restriction to also cover ARC-signing.
389 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
390 in defer=pass mode supply a 450 to the initiator. Previously the message
393 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
394 tls_require_ciphers is used as before.
396 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
397 "pass_unscanned". Only tmpfails from the scanner are written to
398 the paniclog, as they may require admin intervention (permission
399 denied, license issues). Other scanner errors (like decompression
400 bombs) do not cause a paniclog entry.
402 JH/36 Fix reinitialisation of DKIM logging variable between messages.
403 Previously it was possible to log spurious information in receive log
406 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
407 triggered odd behaviour from Outlook Express clients.
409 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
412 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
413 since the IETF WG has not yet settled on that versus the original
414 "bare" representation.
416 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
417 Previously the millisecond value corrupted the output.
418 Fix also for syslog_pid=no and log_selector +pid, for which the pid
419 corrupted the output.
425 JH/01 Rework error string handling in TLS interface so that the caller in
426 more cases is responsible for logging. This permits library-sourced
427 string to be attached to addresses during delivery, and collapses
428 pairs of long lines into single ones.
430 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
431 during configuration. Wildcards are allowed and expanded.
433 JH/02 Rework error string handling in DKIM to pass more info back to callers.
434 This permits better logging.
436 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
437 do not close it down and have the child transport start it up again on
438 the passed-on TCP connection. Instead, proxy the child (and any
439 subsequent ones) for TLS via a unix-domain socket channel. Logging is
440 affected: the continued delivery log lines do not have any DNSSEC, TLS
441 Certificate or OCSP information. TLS cipher information is still logged.
443 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
444 identical IP addresses on different listening ports. Will also affect
447 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
448 add noisy ifdef guards to special-case this sillyness.
449 Patch from Bernd Kuhls.
451 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
452 than 255 are no longer allowed.
454 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
455 Disable the session-cache too, which might reduce our load. Since we
456 currrectly use a new context for every connection, both as server and
457 client, there is no benefit for these.
458 GnuTLS appears to not support tickets server-side by default (we don't
459 call gnutls_session_ticket_enable_server()) but client side is enabled
460 by default on recent versions (3.1.3 +) unless the PFS priority string
463 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
464 <https://reproducible-builds.org/specs/source-date-epoch/>.
466 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
467 the check for any unsuccessful recipients did not notice the limit, and
468 erroneously found still-pending ones.
470 JH/08 Pipeline CHUNKING command and data together, on kernels that support
471 MSG_MORE. Only in-clear (not on TLS connections).
473 JH/09 Avoid using a temporary file during transport using dkim. Unless a
474 transport-filter is involved we can buffer the headers in memory for
475 creating the signature, and read the spool data file once for the
476 signature and again for transmission.
478 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
479 4.77 as the kernel support then wasn't solid, having issues in 64bit
480 mode. Now, it's been long enough. Add support for FreeBSD also.
482 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
483 case where the routing stage had gathered several addresses to send to
484 a host before calling the transport for the first, we previously failed
485 to close down TLS in the old transport process before passing the TCP
486 connection to the new process. The new one sent a STARTTLS command
487 which naturally failed, giving a failed delivery and bloating the retry
488 database. Investigation and fix prototype from Wolfgang Breyha.
490 JH/12 Fix check on SMTP command input synchronisation. Previously there were
491 false-negatives in the check that the sender had not preempted a response
492 or prompt from Exim (running as a server), due to that code's lack of
493 awareness of the SMTP input buffering.
495 PP/04 Add commandline_checks_require_admin option.
496 Exim drops privileges sanely, various checks such as -be aren't a
497 security problem, as long as you trust local users with access to their
498 own account. When invoked by services which pass untrusted data to
499 Exim, this might be an issue. Set this option in main configuration
500 AND make fixes to the calling application, such as using `--` to stop
503 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
504 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
505 responses to those, into a single TLS record each way (this usually means
506 a single packet). As a side issue, smtp_enforce_sync now works on TLS
509 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
510 affects you only if you're dancing at the edge of the param size limits.
511 If you are, and this message makes sense to you, then: raise the
512 configured limit or use OpenSSL 1.1. Nothing we can do for older
515 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
516 cmdline element to get the documented default one. Previously it was
519 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
520 are unsuable for non-message-specific reasons - eg. connection timeout,
521 banner-time rejection.
523 JH/16 Fix logging of delivery remote port, when specified by router, under
526 PP/06 Repair manualroute's ability to take options in any order, even if one
527 is the name of a transport.
530 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
532 JH/17 Change the list-building routines interface to use the expanding-string
533 triplet model, for better allocation and copying behaviour.
535 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
536 Previously it was constructed the first time a possibly-matching string
537 was met in the configuration file input during startup; now it is done
540 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
541 compatible one, to avoid the (poorly documented) possibility of a config
542 file in the working directory redirecting the DB files, possibly correpting
543 some existing file. CVE-2017-10140 assigned for BDB.
545 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
546 cache-hot. Previously, although the result was properly cached, the
547 initial verify call returned a defer.
549 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
550 the main verify for receipient in uncached-mode.
552 JH/22 Retire historical build files to an "unsupported" subdir. These are
553 defined as "ones for which we have no current evidence of testing".
555 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
556 if present. Previously it was ignored.
558 JH/24 Start using specified-initialisers in C structure init coding. This is
559 a C99 feature (it's 2017, so now considered safe).
561 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
562 if was a fixed-sized field and bitmask ops via macros; it is now more
565 PP/07 GitHub PR 56: Apply MariaDB build fix.
566 Patch provided by Jaroslav Å karvada.
568 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
569 during Coverity cleanups [4.87 JH/47]
570 Diagnosis and fix provided by Michael Fischer v. Mollard.
572 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
573 the right size to place the terminating semicolon on its own folded
574 line, the header hash was calculated to an incorrect value thanks to
575 the (relaxed) space the fold became.
577 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
578 and confused the parent.
580 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
581 which could crash as a result. This could lead to undeliverable messages.
583 JH/28 Logging: "next input sent too soon" now shows where input was truncated
586 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
587 matters on fast-turnover and PID-randomising systems, which were getting
588 out-of-order delivery.
590 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
591 a possibly-overlapping copy. The symptom was that "Remote host closed
592 connection in response to HELO" was logged instead of the actual 4xx
595 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
596 Previously only that bufferd was discarded, resulting in SYMTP command
599 JH/32 DKIM: when a message has multiple signatures matching an identity given
600 in dkim_verify_signers, run the dkim acl once for each. Previously only
601 one run was done. Bug 2189.
603 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
604 "panic the current process" to "deliberately defer". The panic log is
605 still written with the problem list name; the mail and reject logs now
606 get a temp-reject line for the message that was being handled, saying
607 something like "domains check lookup or other defer". The SMTP 451
608 message is still "Temporary local problem".
610 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
611 A crafted sequence of BDAT commands could result in in-use memory beeing
612 freed. CVE-2017-16943.
614 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
615 from SMTP input. Previously it was always done; now only done for DATA
616 and not BDAT commands. CVE-2017-16944.
618 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
619 to the message (such as an overlong header line). Previously this was
620 not done and we did not exit BDAT mode. Followon from the previous item
621 though a different problem.
627 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
628 than -2003 did; needs libidn2 in addition to libidn.
630 JH/02 The path option on a pipe transport is now expanded before use.
632 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
633 Patch provided by "Björn", documentation fix added too.
635 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
636 missing a wire-to-host endian conversion.
638 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
639 close after a BDAT command line could be taken as a following command,
640 giving a synch failure. Fix by only checking for synch immediately
641 before acknowledging the chunk.
643 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
644 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
645 macro. Patches provided by Josh Soref.
647 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
648 Previously we did not; the RFC seems ambiguous and VRFY is not listed
649 by IANA as a service extension. However, John Klensin suggests that we
652 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
653 the dkim code may be unix-mode line endings rather than smtp wire-format
654 CRLF, so prepend a CR to any bare LF.
656 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
657 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
659 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
660 the first were themselves being wrongly included in the feed into dkim
661 processing; with most chunk sizes in use this resulted in an incorrect
662 body hash calculated value.
664 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
665 DKIM signature block, for verification. Although advised against by
666 standards it is specifically not ruled illegal.
668 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
670 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
671 missing a body hash (the bh= tag).
673 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
674 It seems that HAProxy sends the Proxy Protocol information in clear and
675 only then does a TLS startup, so do the same.
677 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
678 TCP connections (such as for Spamd) unless the daemon successfully set
679 Fast Open mode on its listening sockets. This fixes breakage seen on
680 too-old kernels or those not configured for Fast Open, at the cost of
681 requiring both directions being enabled for TFO, and TFO never being used
682 by non-daemon-related Exim processes.
684 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
685 endings, at least on the first header line. Try to canonify any that get
686 past that check, despite the cost.
688 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
689 now limited to an arbitrary five deep, while parsing addresses with the
690 strip_excess_angle_brackets option enabled.
692 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
693 instead leave the unprompted TLS handshake in socket buffer for the
694 TLS library to consume.
696 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
698 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
700 JH/16 Drop variables when they go out of scope. Memory management drops a whole
701 region in one operation, for speed, and this leaves assigned pointers
702 dangling. Add checks run only under the testsuite which checks all
703 variables at a store-reset and panics on a dangling pointer; add code
704 explicitly nulling out all the variables discovered. Fixes one known
705 bug: a transport crash, where a dangling pointer for $sending_ip_address
706 originally assigned in a verify callout, is re-used.
708 PP/06 Drop '.' from @INC in various Perl scripts.
710 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
712 PP/08 Reduce a number of compilation warnings under clang; building with
713 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
714 should be warning-free.
716 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
718 HS/01 Fix portability problems introduced by PP/08 for platforms where
719 realloc(NULL) is not equivalent to malloc() [SunOS et al].
721 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
722 chunk. This allows us to accept broken chunked messages. We need a more
723 general solution here.
725 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
726 already-broken messages in the queue.
728 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
730 JH/19 Fix reference counting bug in routing-generated-address tracking.
736 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
737 supports it and a size is available (ie. the sending peer gave us one).
739 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
740 years of being deprecated). The replacements are the ACLs
741 acl_smtp_mime and acl_not_smtp_mime.
743 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
744 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
745 or even an in-clear connection were permitted. Now, if the host lookup
746 was dnssec and dane was requested then the host is only used if the
747 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
748 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
749 if one fails this test.
750 This means that a poorly-configured remote DNS will make it incommunicado;
751 but it protects against a DNS-interception attack on it.
753 JH/04 Bug 1810: make continued-use of an open smtp transport connection
754 non-noisy when a race steals the message being considered.
756 JH/05 If main configuration option tls_certificate is unset, generate a
757 self-signed certificate for inbound TLS connections.
759 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
760 in rewrites and routers.
762 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
763 and logged a warning sing 4.83; now they are a configuration file error.
765 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
766 (lacking @domain). Apply the same qualification processing as RCPT.
768 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
770 JH/10 Support ${sha256:} applied to a string (as well as the previous
773 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
774 a cutthrough deliver is pending, as we always want to make a connection.
775 This also avoids re-routing the message when later placing the cutthrough
776 connection after a verify cache hit.
777 Do not update it with the verify result either.
779 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
780 when routing results in more than one destination address.
782 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
783 signing (which inhibits the cutthrough capability). Previously only
784 the presence of an option was tested; now an expansion evaluating as
785 empty is permissible (obviously it should depend only on data available
786 when the cutthrough connection is made).
788 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
789 the relevant preceding SMTP command did not note the pipelining mode.
791 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
792 Previously they were not counted.
794 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
795 as one having no matching records. Previously we deferred the message
796 that needed the lookup.
798 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
799 distinguished as "(=".
801 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
802 for missing MX records. Previously it only worked for missing A records.
804 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
806 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
807 after the data-go-ahead and data-ack. Patch from Jason Betts.
809 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
810 even for a "none" policy. Patch from Tony Meyer.
812 JH/22 Fix continued use of a connection for further deliveries. If a port was
813 specified by a router, it must also match for the delivery to be
816 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
817 When one of the recipients of a message was unsuitable for the connection
818 (has no matching addresses), we lost track of needing to mark it
819 deferred. As a result mail would be lost.
821 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
823 JH/25 Decoding ACL controls is now done using a binary search; the source code
824 takes up less space and should be simpler to maintain. Merge the ACL
825 condition decode tables also, with similar effect.
827 JH/26 Fix problem with one_time used on a redirect router which returned the
828 parent address unchanged. A retry would see the parent address marked as
829 delivered, so not attempt the (identical) child. As a result mail would
832 JH/27 Fix a possible security hole, wherein a process operating with the Exim
833 UID can gain a root shell. Credit to http://www.halfdog.net/ for
834 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
837 JH/28 Enable {spool,log} filesystem space and inode checks as default.
838 Main config options check_{log,spool}_{inodes,space} are now
839 100 inodes, 10MB unless set otherwise in the configuration.
841 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
842 Previously it only applied to the main-section connection policy
845 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
847 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
848 by me. Added RFC7919 DH primes as an alternative.
850 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
851 are not in the system include path.
853 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
854 GnuTLS, when a session startup failed (eg because the client disconnected)
855 Exim did stdio operations after fclose. This was exposed by a recent
856 change which nulled out the file handle after the fclose.
858 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
859 signed directly by the cert-signing cert, rather than an intermediate
860 OCSP-signing cert. This is the model used by LetsEncrypt.
862 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
864 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
865 an incoming connection.
867 HS/02 Bug 1802: Do not half-close the connection after sending a request
870 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
871 fallback to "prime256v1".
873 JH/34 SECURITY: Use proper copy of DATA command in error message.
874 Could leak key material. Remotely exploitable. CVE-2016-9963.
880 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
881 and 3.4.4 - once the server is enabled to respond to an OCSP request
882 it does even when not requested, resulting in a stapling non-aware
883 client dropping the TLS connection.
885 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
886 support variable-length bit vectors. No functional change.
888 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
889 The I= interface field on outgoing lines is now after the H= remote
890 host field, same as incoming lines. There is a separate
891 outgoing_interface log selector which allows you to disable the
894 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
895 If not running log_selector +smtp_connection the mainlog would be held
896 open indefinitely after a "too many connections" event, including to a
897 deleted file after a log rotate. Leave the per net connection logging
898 leaving it open for efficiency as that will be quickly detected by the
899 check on the next write.
901 HS/01 Bug 1671: Fix post transport crash.
902 Processing the wait-<transport> messages could crash the delivery
903 process if the message IDs didn't exist for some reason. When
904 using 'split_spool_directory=yes' the construction of the spool
905 file name failed already, exposing the same netto behaviour.
907 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
908 mime_regex ACL conditions.
910 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
911 to DSN fail messages (bounces): remote IP, remote greeting, remote response
912 to HELO, local diagnostic string.
914 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
915 log line to debug. Even when configured with a tls authenticator many
916 client connections are expected to not authenticate in this way, so
917 an authenticate fail is not an error.
919 HS/02 Add the Exim version string to the process info. This way exiwhat
920 gives some more detail about the running daemon.
922 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
923 matter for fast-change records such as DNSBLs.
925 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
926 retry record, even if constant. There may be multiple transports with
927 different interface settings and the retry behaviour needs to be kept
930 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
932 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
934 JH/10 Bug 840: fix log_defer_output option of pipe transport
936 JH/11 Bug 830: use same host for all RCPTS of a message, even under
937 hosts_randomize. This matters a lot when combined with mua_wrapper.
939 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
940 ${quote_pgsql:<string>} operator.
942 JH/13 Bug 1708: avoid misaligned access in cached lookup.
944 JH/14 Change header file name for freeradius-client. Relevant if compiling
945 with Radius support; from the Gentoo tree and checked under Fedora.
947 JH/15 Bug 1712: Introduce $prdr_requested flag variable
949 JH/16 Bug 1714: Permit an empty string as expansion result for transport
950 option transport_filter, meaning no filtering.
952 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
954 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
955 defaults to "*" (all hosts). The variable is now available when not built
956 with TLS, default unset, mainly to enable keeping the testsuite sane.
957 If a server certificate is not supplied (via tls_certificate) an error is
958 logged, and clients will find TLS connections fail on startup. Presumably
959 they will retry in-clear.
960 Packagers of Exim are strongly encouraged to create a server certificate
961 at installation time.
963 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
964 with the $config_file variable.
966 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
967 in transport context, after the attempt, and per-recipient. The latter type
968 is per host attempted. The event data is the error message, and the errno
969 information encodes the lookup type (A vs. MX) used for the (first) host,
970 and the trailing two digits of the smtp 4xx response.
972 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
973 to write to mainlog (or rejectlog, paniclog) in the window between file
974 creation and permissions/ownership being changed. Particularly affects
975 installations where exicyclog is run as root, rather than exim user;
976 result is that the running daemon panics and dies.
978 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
980 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
981 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
982 "pri" and "weight". Note that the previous implicit priority given by the
983 list order is no longer honoured.
985 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
988 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
989 by defining SUPPORT_SOCKS.
991 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
992 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
993 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
994 variables are renamed to proxy_{local,external}_{address,port}.
996 JH/27 Move Internationalisation support from Experimental to mainline, enabled
997 for a build by defining SUPPORT_I18N
999 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1000 of the query string, and make ${quote_redis:} do that quoting.
1002 JH/29 Move Events support from Experimental to mainline, enabled by default
1003 and removable for a build by defining DISABLE_EVENT.
1005 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1007 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1008 cached by the daemon.
1010 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1011 by defining LOOKUP_REDIS. The libhiredis library is required.
1013 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1014 keys are given for lookup.
1016 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1017 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1018 only supported when built with TLS support. The PolarSSL SHA routines
1019 are still used when the TLS library is too old for convenient support.
1021 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1022 openssl_options), for security. OpenSSL forces this from version 1.1.0
1023 server-side so match that on older versions.
1025 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1026 allocation for $value could be released as the expansion processing
1027 concluded, but leaving the global pointer active for it.
1029 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1030 and to use the domains and local_parts ACL conditions.
1032 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1033 incorrectly not doubled on cutthrough transmission, hence seen as a
1034 body-termination at the receiving system - resulting in truncated mails.
1035 Commonly the sender saw a TCP-level error, and retransmitted the message
1036 via the normal store-and-forward channel. This could result in duplicates
1037 received - but deduplicating mailstores were liable to retain only the
1038 initial truncated version.
1040 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1042 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1044 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1045 we're in there, support oversigning also; bug 1309.
1047 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1049 HS/04 Add support for keep_environment and add_environment options.
1051 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1052 either intentional arithmetic overflow during PRNG, or testing config-
1055 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1056 delivery resulted in actual delivery. Cancel cutthrough before DATA
1059 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1060 rejects a recipient: pass the reject to the originator.
1062 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1063 Many were false-positives and ignorable, but it's worth fixing the
1066 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1067 for the new environment-manipulation done at startup. Move the routines
1068 from being local to tls.c to being global via the os.c file.
1070 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1071 an extract embedded as result-arg for a map, the first arg for extract
1072 is unavailable so we cannot tell if this is a numbered or keyed
1073 extraction. Accept either.
1079 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1082 JH/02 The smtp transport option "multi_domain" is now expanded.
1084 JH/03 The smtp transport now requests PRDR by default, if the server offers
1087 JH/04 Certificate name checking on server certificates, when exim is a client,
1088 is now done by default. The transport option tls_verify_cert_hostnames
1089 can be used to disable this per-host. The build option
1090 EXPERIMENTAL_CERTNAMES is withdrawn.
1092 JH/05 The value of the tls_verify_certificates smtp transport and main options
1093 default to the word "system" to access the system default CA bundle.
1094 For GnuTLS, only version 3.0.20 or later.
1096 JH/06 Verification of the server certificate for a TLS connection is now tried
1097 (but not required) by default. The verification status is now logged by
1098 default, for both outbound TLS and client-certificate supplying inbound
1101 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1104 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1105 Status Notification (bounce) messages are now MIME format per RFC 3464.
1106 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1107 under the control of the dsn_advertise_hosts option, and routers may
1108 have a dsn_lasthop option.
1110 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1111 default, modifiable by a malware= option. The list separator for
1112 the options can now be changed in the usual way. Bug 68.
1114 JH/10 The smtp_receive_timeout main option is now expanded before use.
1116 JH/11 The incoming_interface log option now also enables logging of the
1117 local interface on delivery outgoing connections.
1119 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1120 if the interface and destination host and port all match.
1122 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1125 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1126 Patch from Andrew Lewis.
1128 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1129 now supports optional time-restrictions, weighting, and priority
1130 modifiers per server. Patch originally by <rommer@active.by>.
1132 JH/16 The spamd_address main option now supports a mixed list of local
1133 and remote servers. Remote servers can be IPv6 addresses, and
1134 specify a port-range.
1136 JH/17 Bug 68: The spamd_address main option now supports an optional
1137 timeout value per server.
1139 JH/18 Bug 1581: Router and transport options headers_add/remove can
1140 now have the list separator specified.
1142 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1145 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1148 JH/21 Support for the A6 type of dns record is withdrawn.
1150 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1151 rather than the verbs used.
1153 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1154 from 255 to 1024 chars.
1156 JH/24 Verification callouts now attempt to use TLS by default.
1158 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1159 are generic router options now. The defaults didn't change.
1161 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1162 Original patch from Alexander Shikoff, worked over by JH.
1164 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1165 files if ZCAT_COMMAND is not executable.
1167 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1169 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1171 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1172 Normally benign, it bites when the pair was led to by a CNAME;
1173 modern usage is to not canonicalize the domain to a CNAME target
1174 (and we were inconsistent anyway for A-only vs AAAA+A).
1176 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1178 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1179 when evaluating $sender_host_dnssec.
1181 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1182 $sender_helo_dnssec variable.
1184 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1186 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1188 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1190 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1191 documented as working, but never had. Support all but $spam_report.
1193 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1194 added for tls authenticator.
1196 HS/03 Add perl_taintmode main config option
1202 TL/01 When running the test suite, the README says that variables such as
1203 no_msglog_check are global and can be placed anywhere in a specific
1204 test's script, however it was observed that placement needed to be near
1205 the beginning for it to behave that way. Changed the runtest perl
1206 script to read through the entire script once to detect and set these
1207 variables, reset to the beginning of the script, and then run through
1208 the script parsing/test process like normal.
1210 TL/02 The BSD's have an arc4random API. One of the functions to induce
1211 adding randomness was arc4random_stir(), but it has been removed in
1212 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1213 function when detected.
1215 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1216 cause callback expansion.
1218 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1219 syntax errors in an expansion can be treated as a string instead of
1220 logging or causing an error, due to the internal use of bool_lax
1221 instead of bool when processing it.
1223 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1224 server certificates when making smtp deliveries.
1226 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1228 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1230 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1232 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1233 Merged patch from Sebastian Wiedenroth.
1235 JH/05 Fix results-pipe from transport process. Several recipients, combined
1236 with certificate use, exposed issues where response data items split
1237 over buffer boundaries were not parsed properly. This eventually
1238 resulted in duplicates being sent. This issue only became common enough
1239 to notice due to the introduction of connection certificate information,
1240 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1242 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1243 size buffer was used, resulting in syntax errors when an expansion
1246 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1247 version 3.3.6 or later.
1249 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1250 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1251 both become "event_action", the variables become $event_name, $event_data
1252 and $event_defer_errno. There is a new variable $verify_mode, usable in
1253 routers, transports and related events. The tls:cert event is now also
1254 raised for inbound connections, if the main configuration event_action
1257 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1258 early OCSP support, but no stapling (appears to be less than 1.0.0).
1260 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1261 server certificate names available under the smtp transport option
1262 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1265 JH/10 Time-related extraction expansions from certificates now use the main
1266 option "timezone" setting for output formatting, and are consistent
1267 between OpenSSL and GnuTLS compilations. Bug 1541.
1269 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1270 encoded parameter in the incoming message. Bug 1558.
1272 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1273 include certificate info, eximon was claiming there were spoolfile
1276 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1278 JH/14 Log delivery-related information more consistently, using the sequence
1279 "H=<name> [<ip>]" wherever possible.
1281 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1282 are problematic for Debian distribution, omit them from the release
1285 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1287 JH/16 Fix string representation of time values on 64bit time_t architectures.
1290 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1291 output list separator was used.
1296 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1297 checkers that were complaining about end of non-void function with no
1300 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1301 This was a regression introduced in 4.83 by another bugfix.
1303 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1305 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1306 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1312 TF/01 Correctly close the server side of TLS when forking for delivery.
1314 When a message was received over SMTP with TLS, Exim failed to clear up
1315 the incoming connection properly after forking off the child process to
1316 deliver the message. In some situations the subsequent outgoing
1317 delivery connection happened to have the same fd number as the incoming
1318 connection previously had. Exim would try to use TLS and fail, logging
1319 a "Bad file descriptor" error.
1321 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1322 utilities have not been installed.
1324 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1325 temporary space as the ACL may create new global variables.
1327 TL/01 LDAP support uses per connection or global context settings, depending
1328 upon the detected version of the libraries at build time.
1330 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1331 to extract and use the src ip:port in logging and expansions as if it
1332 were a direct connection from the outside internet. PPv2 support was
1333 updated based on HAProxy spec change in May 2014.
1335 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1337 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1338 Properly escape header and check for NULL return.
1340 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1343 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1345 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1346 characters in header names, implemented as a verify condition.
1347 Contributed by Michael Fischer v. Mollard.
1349 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1350 results permerror and temperror. Previous values are deprecated but
1351 still accepted. In a future release, err_perm and err_temp will be
1352 completely removed, which will be a backward incompatibility if the
1353 ACL tests for either of these two old results. Patch contributed by
1354 user bes-internal on the mailing list.
1356 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1358 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1359 selectors, in both main and reject logs.
1361 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1364 JH/07 Add malware type "sock" for talking to simple daemon.
1366 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1368 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1369 routers/transports under cutthrough routing.
1371 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1372 numbers. Touch up "bool" conditional to keep the same definition.
1374 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1376 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1378 JH/12 Expand items in router/transport headers_add or headers_remove lists
1379 individually rather than the list as a whole. Bug 1452.
1381 Required for reasonable handling of multiple headers_ options when
1382 they may be empty; requires that headers_remove items with embedded
1383 colons must have them doubled (or the list-separator changed).
1385 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1386 view the policy declared in the DMARC record. Currently, $dmarc_status
1387 is a combined value of both the record presence and the result of the
1390 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1392 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1393 dnslookup router and the smtp transport (applying to the forward
1396 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1397 of ldap servers used for a specific lookup. Patch provided by Heiko
1400 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1401 New variable $lookup_dnssec_authenticated for observability.
1403 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1404 Patch submitted by Lars Timman.
1406 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1408 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1409 Requires trusted mode and valid format message id, aborts otherwise.
1410 Patch contributed by Heiko Schlichting.
1412 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1413 certextract with support for various fields. Bug 1358.
1415 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1416 is requested by default, modifiable by smtp transport option
1419 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1420 operate on certificate variables to give certificate fingerprints
1421 Also new ${sha256:cert_variable}.
1423 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1425 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1426 Christian Aistleitner.
1428 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1430 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1431 file. Patch from Wolfgang Breyha.
1433 JH/25 Expand the coverage of the delivery $host and $host_address to
1434 client authenticators run in verify callout. Bug 1476.
1436 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1437 align with daemon_smtp_ports. Bug 72.
1439 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1440 support and error reporting did not work properly.
1442 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1443 and is readable. Patch from Andrew Colin Kissa.
1445 TL/14 Enhance documentation of ${run expansion and how it parses the
1446 commandline after expansion, particularly in the case when an
1447 unquoted variable expansion results in an empty value.
1449 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1451 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1452 by renaming away from it. Addresses GH issue 15, reported by
1455 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1457 TL/15 SECURITY: prevent double expansion in math comparison functions
1458 (can expand unsanitized data). Not remotely exploitable.
1465 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1467 PP/02 Make -n do something, by making it not do something.
1468 When combined with -bP, the name of an option is not output.
1470 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1473 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1474 $sender_host_name and config options to manage this, and basic check
1477 PP/05 DSCP support for outbound connections and control modifier for inbound.
1479 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1480 (Only plugin which currently uses this is kerberos4, which nobody should
1481 be using, but we should make it available and other future plugins might
1482 conceivably use it, even though it would break NAT; stuff *should* be
1483 using channel bindings instead).
1485 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1486 name; added for Sendmail compatibility; requires admin caller.
1487 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1488 just ignore it); requires trusted caller.
1489 Also parse but ignore: -Ac -Am -X<logfile>
1492 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1494 TL/02 Add +smtp_confirmation as a default logging option.
1496 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1497 Patch by Magnus Holmgren from 2007-02-20.
1499 TL/04 Bugzilla 1281 - Spec typo.
1500 Bugzilla 1283 - Spec typo.
1501 Bugzilla 1290 - Spec grammar fixes.
1503 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1505 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1507 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1508 mailing list by by Dmitry Isaikin.
1510 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1512 JH/02 Support "G" suffix to numbers in ${if comparisons.
1514 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1516 NM/01 Bugzilla 1197 - Spec typo
1517 Bugzilla 1196 - Spec examples corrections
1519 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1521 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1522 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1525 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1526 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1528 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1529 "acl {{name}{arg}...}", and optional args on acl condition
1532 JH/05 Permit multiple router/transport headers_add/remove lines.
1534 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1536 JH/07 Avoid using a waiting database for a single-message-only transport.
1537 Performance patch from Paul Fisher. Bugzilla 1262.
1539 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1542 JH/09 Add $headers_added variable, with content from use of ACL modifier
1543 add_header (but not yet added to the message). Bugzilla 199.
1545 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1546 Pulled from Bugzilla 817 by Wolfgang Breyha.
1548 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1550 (nb: this is the same fix as in Exim 4.80.1)
1552 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1555 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1558 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1560 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1561 advertise SMTP AUTH mechanism to us, instead of a generic
1562 protocol violation error. Also, make Exim more robust to bad
1563 data from the Dovecot auth socket.
1565 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1567 When a queue runner is handling a message, Exim first routes the
1568 recipient addresses, during which it prunes them based on the retry
1569 hints database. After that it attempts to deliver the message to
1570 any remaining recipients. It then updates the hints database using
1573 So if a recipient address works intermittently, it can get repeatedly
1574 deferred at routing time. The retry hints record remains fresh so the
1575 address never reaches the final cutoff time.
1577 This is a fairly common occurrence when a user is bumping up against
1578 their storage quota. Exim had some logic in its local delivery code
1579 to deal with this. However it did not apply to per-recipient defers
1580 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1582 This change adds a proper retry rule check during routing so that the
1583 final cutoff time is checked against the message's age. We only do
1584 this check if there is an address retry record and there is not a
1585 domain retry record; this implies that previous attempts to handle
1586 the address had the retry_use_local_parts option turned on. We use
1587 this as an approximation for the destination being like a local
1588 delivery, as in LMTP.
1590 I suspect this new check makes the old local delivery cutoff check
1591 redundant, but I have not verified this so I left the code in place.
1593 TF/02 Correct gecos expansion when From: is a prefix of the username.
1595 Test 0254 submits a message to Exim with the header
1599 When I ran the test suite under the user fanf2, Exim expanded
1600 the header to contain my full name, whereas it should have added
1601 a Resent-Sender: header. It erroneously treats any prefix of the
1602 username as equal to the username.
1604 This change corrects that bug.
1606 GF/01 DCC debug and logging tidyup
1607 Error conditions log to paniclog rather than rejectlog.
1608 Debug lines prefixed by "DCC: " to remove any ambiguity.
1610 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1612 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1613 Bug spotted by Jeremy Harris; was flawed since initial commit.
1614 Would have resulted in OCSP responses post-SNI triggering an Exim
1615 NULL dereference and crash.
1617 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1619 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1620 Bug detection, analysis and fix by Samuel Thibault.
1621 Bugzilla 1331, Debian bug #698092.
1623 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1625 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1626 Server implementation by Todd Lyons, client by JH.
1627 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1628 config variable "prdr_enable" controls whether the server
1629 advertises the facility. If the client requests PRDR a new
1630 acl_data_smtp_prdr ACL is called once for each recipient, after
1631 the body content is received and before the acl_smtp_data ACL.
1632 The client is controlled by both of: a hosts_try_prdr option
1633 on the smtp transport, and the server advertisement.
1634 Default client logging of deliveries and rejections involving
1635 PRDR are flagged with the string "PRDR".
1637 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1638 fclose(). Diagnosis by Todd Lyons.
1640 PP/17 Update configure.default to handle IPv6 localhost better.
1641 Patch by Alain Williams (plus minor tweaks).
1644 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1645 This is now consistent with GnuTLS, and is now documented: the
1646 previous undocumented portable approach to treating the option as
1647 unset was to force an expansion failure. That still works, and
1648 an empty string is now equivalent.
1650 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1651 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1652 not performing validation itself.
1654 PP/20 Added force_command boolean option to pipe transport.
1655 Patch from Nick Koston, of cPanel Inc.
1657 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1660 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1662 PP/21 Fix eximon continuous updating with timestamped log-files.
1663 Broken in a format-string cleanup in 4.80, missed when I repaired the
1664 other false fix of the same issue.
1665 Report and fix from Heiko Schlichting.
1668 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1669 Report from Prashanth Katuri.
1671 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1672 It's SecureTransport, so affects any MacOS clients which use the
1673 system-integrated TLS libraries, including email clients.
1675 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1676 using a MIME ACL for non-SMTP local injection.
1677 Report and assistance in diagnosis by Warren Baker.
1679 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1681 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1683 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1684 last id that failed so it may be referenced in subsequent ACL's.
1686 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1689 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1690 ldap library initialization, allowing self-signed CA's to be
1691 used. Also properly sets require_cert option later in code by
1692 using NULL (global ldap config) instead of ldap handle (per
1693 session). Bug diagnosis and testing by alxgomz.
1695 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1696 the src/util/ subdirectory.
1698 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1699 renamed to Transport Post Delivery Action by Jeremy Harris, as
1702 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1703 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1704 redis_servers = needs to be configured which will be used by the redis
1705 lookup. Patch from Warren Baker, of The Packet Hub.
1707 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1709 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1710 hostname or reverse DNS when processing a host list. Used suggestions
1711 from multiple comments on this bug.
1713 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1715 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1716 Missed a few lines, added it to make the runtest require no keyboard
1719 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1720 contains upper case chars. Make router use caseful_local_part.
1722 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1723 support when GnuTLS has been built with p11-kit.
1729 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1731 This, or similar/improved, will also be change PP/11 of 4.82.
1737 PP/01 Handle short writes when writing local log-files.
1738 In practice, only affects FreeBSD (8 onwards).
1739 Bugzilla 1053, with thanks to Dmitry Isaikin.
1741 NM/01 Bugzilla 949 - Documentation tweak
1743 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1746 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1748 PP/02 Implemented gsasl authenticator.
1750 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1752 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1753 `pkg-config foo` for cflags/libs.
1755 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1756 with rest of GSASL and with heimdal_gssapi.
1758 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1759 `pkg-config foo` for cflags/libs for the TLS implementation.
1761 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1762 properties get this fed in as external SSF. A number of robustness
1763 and debugging improvements to the cyrus_sasl authenticator.
1765 PP/08 cyrus_sasl server now expands the server_realm option.
1767 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1768 Patch by Jeremy Harris.
1770 PP/10 Added dbmjz lookup type.
1772 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1774 PP/12 MAIL args handles TAB as well as SP, for better interop with
1775 non-compliant senders.
1776 Analysis and variant patch by Todd Lyons.
1778 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1779 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1780 Patch from Dirk Mueller <dmueller@suse.com>
1782 PP/13 tls_peerdn now print-escaped for spool files.
1783 Observed some $tls_peerdn in wild which contained \n, which resulted
1784 in spool file corruption.
1786 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1787 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1788 or write after TLS renegotiation, which otherwise led to messages
1791 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1792 as a tracking header (ie: a signed header comes before the signature).
1793 Patch from Wolfgang Breyha.
1795 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1796 comma-sep list; embedded commas doubled.
1798 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1800 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1802 Report and patch from Dmitry Banschikov.
1804 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1805 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1806 failures. We appear to now support TLS1.1+ with Exim.
1808 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1809 lets Exim select keys and certificates based upon TLS SNI from client.
1810 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1811 before an outbound SMTP session. New log_selector, +tls_sni.
1813 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1814 NULL dereference. Report and patch from Alun Jones.
1816 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1817 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1818 Not seeing resolver debug output on NetBSD, but suspect this is a
1819 resolver implementation change.
1821 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1822 Left warnings. Added "eximon gdb" invocation mode.
1824 PP/21 Defaulting "accept_8bitmime" to true, not false.
1826 PP/22 Added -bw for inetd wait mode support.
1828 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1829 locate the relevant includes and libraries. Made this the default.
1831 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1832 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1834 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1835 This may cause build issues on older platforms.
1837 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1838 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1839 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1840 Added SNI support via GnuTLS too.
1841 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1843 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1845 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1846 Applied second patch from Janne, implementing suggestion to default
1847 multiple-strings-in-record handling to match SPF spec.
1849 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1851 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1852 read-only, out of scope).
1853 Patch from Wolfgang Breyha, report from Stuart Northfield.
1855 PP/29 Fix three issues highlighted by clang analyser static analysis.
1856 Only crash-plausible issue would require the Cambridge-specific
1857 iplookup router and a misconfiguration.
1858 Report from Marcin Mirosław.
1860 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1862 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1863 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1864 As part of this, removing so much warning spew let me fix some minor
1865 real issues in debug logging.
1867 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1868 assignment on my part. Fixed.
1870 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1871 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1872 Janne Snabb (who went above and beyond: thank you).
1874 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1875 string otherwise requires a connection and a bunch more work and it's
1876 relatively easy to get wrong. Should also expose TLS library linkage
1879 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1880 64-bit ${eval} (JH/03).
1882 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1883 GNU libc to support some of the 64-bit stuff, should not lead to
1884 conflicts. Defined before os.h is pulled in, so if a given platform
1885 needs to override this, it can.
1887 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1888 protection layer was required, which is not implemented.
1889 Bugzilla 1254, patch from Wolfgang Breyha.
1891 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1892 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1893 tls_dhparam take prime identifiers. Also unbreak combination of
1894 OpenSSL+DH_params+TLSSNI.
1896 PP/39 Disable SSLv2 by default in OpenSSL support.
1902 PP/01 Solaris build fix for Oracle's LDAP libraries.
1903 Bugzilla 1109, patch from Stephen Usher.
1905 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1907 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1910 TF/02 Fix a couple more cases where we did not log the error message
1911 when unlink() failed. See also change 4.74-TF/03.
1913 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1914 lock up or crash if it happened to be inside a call to libc when it
1915 got a SIGUSR1 from exiwhat.
1917 The SIGUSR1 handler appends the current process status to the process
1918 log which is later printed by exiwhat. It used to use the general
1919 purpose logging code to do this, but several functions it calls are
1920 not safe for signals.
1922 The new output code in the SIGUSR1 handler is specific to the process
1923 log, and simple enough that it's easy to inspect for signal safety.
1924 Removing some special cases also simplifies the general logging code.
1925 Removing the spurious timestamps from the process log simplifies
1928 TF/04 Improved ratelimit ACL condition.
1930 The /noupdate option has been deprecated in favour of /readonly which
1931 has clearer semantics. The /leaky, /strict, and /readonly update modes
1932 are mutually exclusive. The update mode is no longer included in the
1933 database key; it just determines when the database is updated. (This
1934 means that when you upgrade Exim will forget old rate measurements.)
1936 Exim now checks that the per_* options are used with an update mode that
1937 makes sense for the current ACL. For example, when Exim is processing a
1938 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1939 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1940 must specify per_mail/readonly. If you omit the update mode it defaults to
1941 /leaky where that makes sense (as before) or /readonly where required.
1943 The /noupdate option is now undocumented but still supported for
1944 backwards compatibility. It is equivalent to /readonly except that in
1945 ACLs where /readonly is required you may specify /leaky/noupdate or
1946 /strict/noupdate which are treated the same as /readonly.
1948 A useful new feature is the /count= option. This is a generalization
1949 of the per_byte option, so that you can measure the throughput of other
1950 aggregate values. For example, the per_byte option is now equivalent
1951 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1953 The per_rcpt option has been generalized using the /count= mechanism
1954 (though it's more complicated than the per_byte equivalence). When it is
1955 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1956 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1957 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1958 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1959 Note that using per_rcpt with a non-readonly update mode in more than
1960 one ACL will cause the recipients to be double-counted. (The per_mail
1961 and per_byte options don't have this problem.)
1963 The handling of very low rates has changed slightly. If the computed rate
1964 is less than the event's count (usually one) then this event is the first
1965 after a long gap. In this case the rate is set to the same as this event's
1966 count, so that the first message of a spam run is counted properly.
1968 The major new feature is a mechanism for counting the rate of unique
1969 events. The new per_addr option counts the number of different
1970 recipients that someone has sent messages to in the last time period. It
1971 behaves like per_rcpt if all the recipient addresses are different, but
1972 duplicate recipient addresses do not increase the measured rate. Like
1973 the /count= option this is a general mechanism, so the per_addr option
1974 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1975 example, measure the rate that a client uses different sender addresses
1976 with the options per_mail/unique=$sender_address. There are further
1977 details in the main documentation.
1979 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1981 TF/06 Removed a few PCRE remnants.
1983 TF/07 Automatically extract Exim's version number from tags in the git
1984 repository when doing development or release builds.
1986 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1987 Bugzilla 879. Patch from Paul Fisher.
1989 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1990 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1993 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1995 PP/05 Variable $av_failed, true if the AV scanner deferred.
1996 Bugzilla 1078. Patch from John Horne.
1998 PP/06 Stop make process more reliably on build failure.
1999 Bugzilla 1087. Patch from Heiko Schlittermann.
2001 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2002 Bugzilla 1089. Patch from Heiko Schlittermann.
2004 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2005 Bugzilla 1131. Patch from Holger Weiß.
2007 PP/09 Handle IPv6 addresses with SPF.
2008 Bugzilla 860. Patch from Wolfgang Breyha.
2010 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2012 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2015 PP/11 match_* no longer expand right-hand-side by default.
2016 New compile-time build option, EXPAND_LISTMATCH_RHS.
2017 New expansion conditions, "inlist", "inlisti".
2019 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2021 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2023 PP/14 fix log_write() format string regression from TF/03.
2024 Bugzilla 1152. Patch from Dmitry Isaikin.
2030 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2032 PP/02 Harmonised TLS library version reporting; only show if debugging.
2033 Layout now matches that introduced for other libraries in 4.74 PP/03.
2035 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2037 PP/04 New "dns_use_edns0" global option.
2039 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2042 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2043 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2045 TK/01 Updated PolarSSL code to 0.14.2.
2046 Bugzilla 1097. Patch from Andreas Metzler.
2048 PP/07 Catch divide-by-zero in ${eval:...}.
2049 Fixes bugzilla 1102.
2051 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2054 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2055 format-string attack -- SECURITY: remote arbitrary code execution.
2057 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2058 time unintentionally subject to list matching rules, letting the header
2059 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2060 arbitrary string expansion). This allowed for information disclosure.
2062 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2063 INT_MIN/-1 -- value coerced to INT_MAX.
2069 NM/01 Workaround for PCRE version dependency in version reporting
2072 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2073 This fixes portability to compilers other than gcc, notably
2074 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2076 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2077 makefiles for portability to HP-UX and POSIX correctness.
2079 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2080 Also via indented variable definition in the Makefile.
2081 (Debugging by Oliver Heesakkers).
2083 PP/02 Restore caching of spamd results with expanded spamd_address.
2084 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2086 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2087 Improves build reliability. Fix from: Frank Elsner
2089 NM/02 Fix wide character breakage in the rfc2047 coding
2090 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2092 NM/03 Allow underscore in dnslist lookups
2093 Fixes bug 1026. Patch from Graeme Fowler
2095 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2096 Code patches from Adam Ciarcinski of NetBSD.
2098 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2101 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2102 is logged, to avoid truncation. Patch from John Horne.
2104 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2105 Patch from Jakob Hirsch.
2107 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2108 SQL string expansion failure details.
2109 Patch from Andrey Oktyabrski.
2111 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2112 Patch from Simon Arlott.
2114 PP/09 New lookups functionality failed to compile on old gcc which rejects
2115 extern declarations in function scope.
2116 Patch from Oliver Fleischmann
2118 PP/10 Use sig_atomic_t for flags set from signal handlers.
2119 Check getgroups() return and improve debugging.
2120 Fixed developed for diagnosis in bug 927 (which turned out to be
2123 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2124 Patch from Mark Zealey.
2126 PP/12 Bugzilla 1056: Improved spamd server selection.
2127 Patch from Mark Zealey.
2129 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2130 Based on patch from Heiko Schlittermann.
2132 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2133 Patch from Uwe Doering, sign-off by Michael Haardt.
2135 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2136 variable declaration deep within a block. Bug and patch from
2139 PP/15 lookups-Makefile IRIX compatibility coercion.
2141 PP/16 Make DISABLE_DKIM build knob functional.
2143 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2144 Patch by Simon Arlott
2146 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2147 variable argument macros. Our copy now differs from upstream.
2153 TF/01 Failure to get a lock on a hints database can have serious
2154 consequences so log it to the panic log.
2156 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2157 controlled using the smtp_confirmation log selector.
2159 TF/03 Include the error message when we fail to unlink a spool file.
2161 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2162 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2163 for maintaining out-of-tree patches for some time.
2165 PP/01 Bugzilla 139: Documentation and portability issues.
2166 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2167 Handle per-OS dynamic-module compilation flags.
2169 PP/02 Let /dev/null have normal permissions.
2170 The 4.73 fixes were a little too stringent and complained about the
2171 permissions on /dev/null. Exempt it from some checks.
2172 Reported by Andreas M. Kirchwitz.
2174 PP/03 Report version information for many libraries, including
2175 Exim version information for dynamically loaded libraries. Created
2176 version.h, now support a version extension string for distributors
2177 who patch heavily. Dynamic module ABI change.
2179 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2180 privilege escalation vulnerability whereby the Exim run-time user
2181 can cause root to append content of the attacker's choosing to
2184 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2187 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2188 If dropping privileges for untrusted macros, we disabled normal logging
2189 on the basis that it would fail; for the Exim run-time user, this is not
2190 the case, and it resulted in successful deliveries going unlogged.
2191 Fixed. Reported by Andreas Metzler.
2197 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2198 only prepend for the Resent-* case. Fixes regression introduced in
2199 Exim 4.70 by NM/22 for Bugzilla 607.
2201 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2202 increasing numbers of administrators be bitten by this.
2204 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2206 PP/03 Bugzilla 994: added openssl_options main configuration option.
2208 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2210 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2212 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2214 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2215 without a peer certificate, leading to a segfault because of an
2216 assumption that peers always have certificates. Be a little more
2217 paranoid. Problem reported by Martin Tscholak.
2219 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2220 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2221 NB: ClamAV planning to remove STREAM in "middle of 2010".
2222 CL also introduces -bmalware, various -d+acl logging additions and
2223 more caution in buffer sizes.
2225 PP/09 Implemented reverse_ip expansion operator.
2227 PP/10 Bugzilla 937: provide a "debug" ACL control.
2229 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2231 PP/12 Bugzilla 973: Implement --version.
2233 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2235 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2237 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2239 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2240 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2241 ignore trailing whitespace.
2243 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2245 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2248 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2249 Notification from Dr Andrew Aitchison.
2251 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2252 ExtendedDetectionInfo response format.
2253 Notification from John Horne.
2255 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2258 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2259 XSL and documented dependency on system catalogs, with examples of how
2262 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2265 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2266 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2267 configuration file which is writeable by the Exim user or group.
2269 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2270 of configuration files to cover files specified with the -C option if
2271 they are going to be used with root privileges, not just the default
2274 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2275 option (effectively making it always true).
2277 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2278 files to be used while preserving root privileges.
2280 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2281 that rogue child processes cannot use them.
2283 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2284 run-time user, instead of root.
2286 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2287 Exim run-time user without dropping privileges.
2289 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2290 result string, instead of calling string_vformat() twice with the same
2293 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2294 for other users. Others should always drop root privileges if they use
2295 -C on the command line, even for a whitelisted configure file.
2297 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2299 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2305 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2306 $data_path, and $header_path variables; fixed documentation bugs and
2309 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2310 exipick to access non-standard spools, including the "frozen" queue
2313 NM/01 Bugzilla 965: Support mysql stored procedures.
2314 Patch from Alain Williams
2316 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2318 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2319 Patch from Andreas Metzler
2321 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2322 Patch from Kirill Miazine
2324 NM/05 Bugzilla 671: Added umask to procmail example.
2326 JJ/03 installed exipick 20100323.0, fixing doc bug
2328 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2329 directory. Notification and patch from Dan Rosenberg.
2331 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2333 TK/02 Improve log output when DKIM signing operation fails.
2335 MH/01 Treat the transport option dkim_domain as a colon separated
2336 list, not as a single string, and sign the message with each element,
2337 omitting multiple occurences of the same signer.
2339 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2340 Bugzilla 985, 986. Patch by Simon Arlott
2342 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2343 Patch by Simon Arlott
2345 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2346 MBX locking. Notification from Dan Rosenberg.
2352 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2354 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2356 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2358 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2360 NM/04 Bugzilla 915: Flag broken perl installation during build.
2366 TK/01 Added patch by Johannes Berg that expands the main option
2367 "spamd_address" if it starts with a dollar sign.
2369 TK/02 Write list of recipients to X-Envelope-Sender header when building
2370 the mbox-format spool file for content scanning (suggested by Jakob
2373 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2374 (http://www.dcc-servers.net/) support via dccifd. Activated by
2375 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2377 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2378 by Mark Daniel Reidel <mr@df.eu>.
2380 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2381 When building exim an external PCRE library is now needed -
2382 PCRE is a system library on the majority of modern systems.
2383 See entry on PCRE_LIBS in EDITME file.
2385 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2386 conversation. Added nologin parameter to request.
2387 Patch contributed by Kirill Miazine.
2389 TF/01 Do not log submission mode rewrites if they do not change the address.
2391 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2393 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2394 log files in place. Contributed by Roberto Lima.
2396 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2398 TF/03 Bugzilla 615: When checking the local_parts router precondition
2399 after a local_part_suffix or local_part_prefix option, Exim now
2400 does not use the address's named list lookup cache, since this
2401 contains cached lookups for the whole local part.
2403 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2404 Robert Millan. Documentation is in experimental-spec.txt.
2406 TF/04 Bugzilla 668: Fix parallel build (make -j).
2408 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2410 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2411 Patch provided by Jan Srzednicki.
2413 TF/05 Leading white space used to be stripped from $spam_report which
2414 wrecked the formatting. Now it is preserved.
2416 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2417 that they are available at delivery time.
2419 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2421 TF/08 TLS error reporting now respects the incoming_interface and
2422 incoming_port log selectors.
2424 TF/09 Produce a more useful error message if an SMTP transport's hosts
2425 setting expands to an empty string.
2427 NM/06 Bugzilla 744: EXPN did not work under TLS.
2428 Patch provided by Phil Pennock.
2430 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2431 Patch provided by Richard Godbee.
2433 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2434 acl_smtp_notquit, added index entry.
2436 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2437 Patch provided by Eugene Bujak.
2439 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2440 accept(). Patch provided by Maxim Dounin.
2442 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2443 Patch provided by Phil Pennock.
2445 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2447 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2448 Patch provided by Brad "anomie" Jorsch.
2450 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2451 Patch provided by Dean Brooks.
2453 TK/05 Add native DKIM support (does not depend on external libraries).
2455 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2456 Patch provided by Graeme Fowler.
2458 NM/16 Bugzilla 851: Documentation example syntax fix.
2460 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2462 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2465 NM/19 Bugzilla 745: TLS version reporting.
2466 Patch provided by Phil Pennock.
2468 NM/20 Bugzilla 167: bool: condition support.
2469 Patch provided by Phil Pennock.
2471 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2472 clients. Patch provided by Phil Pennock.
2474 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2475 Patch provided by Brad "anomie" Jorsch.
2477 NM/23 Bugzilla 687: Fix misparses in eximstats.
2478 Patch provided by Heiko Schlittermann.
2480 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2481 Patch provided by Heiko Schlittermann.
2483 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2484 plus update to original patch.
2486 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2488 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2489 Patch provided by David Brownlee.
2491 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2493 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2495 NM/30 Bugzilla 888: TLS documentation bugfixes.
2497 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2499 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2500 Unlike the original bugzilla I have changed all shell scripts in src tree.
2502 NM/33 Bugzilla 898: Transport filter timeout fix.
2503 Patch by Todd Rinaldo.
2505 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2506 Patch by Serge Demonchaux.
2508 NM/35 Bugzilla 39: Base64 decode bug fixes.
2509 Patch by Jakob Hirsch.
2511 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2513 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2515 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2517 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2523 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2524 ALT-N's libdkim that I have put here:
2525 http://duncanthrax.net/exim-experimental/
2527 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2528 were called 'true' and it seems that C99 defines that as a reserved
2529 keyword to be used with 'bool' variable types. That means you could
2530 not include C99-style headers which use bools without triggering
2531 build errors in sieve.c.
2533 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2534 as mailq or other aliases. Changed the --help handling significantly
2535 to do whats expected. exim_usage() emits usage/help information.
2537 SC/01 Added the -bylocaldomain option to eximstats.
2539 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2541 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2543 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2549 PH/01 Another patch from the Sieve maintainer.
2551 PH/02 When an IPv6 address is converted to a string for single-key lookup
2552 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2553 dots are used instead of colons so that keys in lsearch files need not
2554 contain colons. This was done some time before quoting was made available
2555 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2556 (notated using the quote facility) so as to distinguish them from IPv4
2557 keys. This meant that lookups for IP addresses in host lists did not work
2558 for iplsearch lookups.
2560 This has been fixed by arranging for IPv6 addresses to be expressed with
2561 colons if the lookup type is iplsearch. This is not incompatible, because
2562 previously such lookups could never work.
2564 The situation is now rather anomalous, since one *can* have colons in
2565 ordinary lsearch keys. However, making the change in all cases is
2566 incompatible and would probably break a number of configurations.
2568 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2571 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2572 conversion specification without a maximum field width, thereby enabling
2573 a rogue spamd server to cause a buffer overflow. While nobody in their
2574 right mind would setup Exim to query an untrusted spamd server, an
2575 attacker that gains access to a server running spamd could potentially
2576 exploit this vulnerability to run arbitrary code as the Exim user.
2578 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2579 $primary_hostname instead of what libspf2 thinks the hosts name is.
2581 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2582 a directory entry by the name of the lookup key. Previously, if a
2583 symlink pointed to a non-existing file or a file in a directory that
2584 Exim lacked permissions to read, a lookup for a key matching that
2585 symlink would fail. Now it is enough that a matching directory entry
2586 exists, symlink or not. (Bugzilla 503.)
2588 PH/03 The body_linecount and body_zerocount variables are now exported in the
2591 PH/04 Added the $dnslist_matched variable.
2593 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2594 This means they are set thereafter only if the connection becomes
2597 PH/06 Added the client_condition to authenticators so that some can be skipped
2598 by clients under certain conditions.
2600 PH/07 The error message for a badly-placed control=no_multiline_responses left
2601 "_responses" off the end of the name.
2603 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2605 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2606 (without spaces) instead of just copying the configuration text.
2608 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2610 PH/11 Added $max_received_linelength.
2612 PH/12 Added +ignore_defer and +include_defer to host lists.
2614 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2615 way in which PCRE > 7.0 is built.
2617 PH/14 Implemented queue_only_load_latch.
2619 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2620 MAIL command. The effect was to mangle the value on 64-bit systems.
2622 PH/16 Another patch from the Sieve maintainer.
2624 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2626 PH/18 If a system quota error occurred while trying to create the file for
2627 a maildir delivery, the message "Mailbox is full" was not appended to the
2628 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2629 only to a quota excession during the actual writing of the file.
2631 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2632 characters?) which causes problems in log lines. The DN values are now
2633 passed through string_printing() before being added to log lines.
2635 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2636 and InterBase are left for another time.)
2638 PH/21 Added message_body_newlines option.
2640 PH/22 Guard against possible overflow in moan_check_errorcopy().
2642 PH/23 POSIX allows open() to be a macro; guard against that.
2644 PH/24 If the recipient of an error message contained an @ in the local part
2645 (suitably quoted, of course), incorrect values were put in $domain and
2646 $local_part during the evaluation of errors_copy.
2652 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2653 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2656 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2657 issue a MAIL command.
2659 PH/02 In an ACL statement such as
2661 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2663 if a client was not listed at all, or was listed with a value other than
2664 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2665 the condition was not true (as it should be), so access was not denied.
2666 The bug was that the ! inversion was incorrectly passed on to the second
2667 item. This has been fixed.
2669 PH/03 Added additional dnslists conditions == and =& which are different from
2670 = and & when the dns lookup returns more than one IP address.
2672 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2673 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2675 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2676 FSYNC, which compiles an option called disable_fsync that allows for
2677 bypassing fsync(). The documentation is heavily laced with warnings.
2679 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2681 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2682 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2683 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2684 including adding "make clean"; (3) Added -fPIC when compiling the test
2685 dynamically loaded module, to get rid of a warning.
2687 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2688 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2689 = 0s. The bug is otherwise harmless.
2691 PH/07 There was a bug in the dovecot authenticator such that the value of
2692 $auth1 could be overwritten, and so not correctly preserved, after a
2693 successful authentication. This usually meant that the value preserved by
2694 the server_setid option was incorrect.
2696 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2698 PH/09 Installed PCRE release 7.0.
2700 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2701 run for batched SMTP input. It is now run at the start of every message
2702 in the batch. While fixing this I discovered that the process information
2703 (output by running exiwhat) was not always getting set for -bs and -bS
2704 input. This is fixed, and it now also says "batched" for BSMTP.
2706 PH/11 Added control=no_pipelining.
2708 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2709 patch, slightly modified), and move the expansion of helo_data till after
2710 the connection is made in the smtp transport (so it can use these
2713 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2715 PH/14 Added log_selector = +pid.
2717 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2719 PH/16 Add ${if forany and ${if forall.
2721 PH/17 Added dsn_from option to vary the From: line in DSNs.
2723 PH/18 Flush SMTP output before performing a callout, unless control =
2724 no_callout_flush is set.
2726 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2727 was true (the default) a successful delivery failed to delete the retry
2728 item, thus causing premature timeout of the address. The bug is now
2731 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2733 PH/21 Long custom messages for fakedefer and fakereject are now split up
2734 into multiline responses in the same way that messages for "deny" and
2735 other ACL rejections are.
2737 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2738 with slight modification.
2740 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2741 draft, changing the syntax and factoring some duplicate code".
2743 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2744 for deliveries of the second and subsequent messages over the same SMTP
2747 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2748 ${reduce, with only minor "tidies".
2750 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2752 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2753 expansion side effects.
2755 PH/27 When a message times out after an over-quota error from an Exim-imposed
2756 quota, the bounce message says "mailbox is full". This message was not
2757 being given when it was a system quota that was exceeded. It now should
2760 MH/03 Made $recipients available in local_scan(). local_scan() already has
2761 better access to the recipient list through recipients_list[], but
2762 $recipients can be useful in postmaster-provided expansion strings.
2764 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2765 in the case of a MAIL command with additional options following the
2766 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2767 were accidentally chopped off.
2769 PH/29 SMTP synchronization checks are implemented when a command is read -
2770 there is a check that no more input is waiting when there shouldn't be
2771 any. However, for some commands, a delay in an ACL can mean that it is
2772 some time before the response is written. In this time, more input might
2773 arrive, invalidly. So now there are extra checks after an ACL has run for
2774 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2775 pipelining has not been advertised.
2777 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2779 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2780 returned in the SMTP response when smtp_return_error_details was false.
2781 This has been fixed.
2783 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2784 instead of the C I/O that was originally supplied, because problems were
2785 reported on Solaris.
2787 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2788 Exim which did not show up earlier: it was assuming that a call to
2789 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2790 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2791 was a macro that became an assignment, so it seemed to work. This has
2792 changed to a proper function call with a void return, hence the compile
2793 error. Exim's code has been fixed.
2795 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2798 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2800 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2802 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2803 because it turns out to be too big or there is a timeout) there is no
2804 "Completed" line in the log. When some messages of this type were
2805 selected by exigrep, they were listed as "not completed". Others were
2806 picked up by some special patterns. I have improved the selection
2807 criteria to be more general.
2809 PH/38 The host_find_failed option in the manualroute router can now be set
2810 to "ignore", to completely ignore a host whose IP address cannot be
2811 found. If all hosts are ignored, the behaviour is controlled by the new
2812 host_all_ignored option.
2814 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2815 homing or because of multiple MX records with /mx) generated more than
2816 one IP address, and the following item turned out to be the local host,
2817 all the secondary addresses of the first item were incorrectly removed
2818 from the list, along with the local host and any following hosts (which
2819 is what is supposed to happen).
2821 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2822 whoever called Exim into the -H file. In the case of the daemon it was
2823 behaving confusingly. When first started, it used values for whoever
2824 started the daemon, but after a SIGHUP it used the Exim user (because it
2825 calls itself on a restart). I have changed the code so that it now always
2828 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2829 message are rejected with the same error (e.g. no authentication or bad
2830 sender address), and a DATA command is nevertheless sent (as can happen
2831 with PIPELINING or a stupid MUA), the error message that was given to the
2832 RCPT commands is included in the rejection of the DATA command. This is
2833 intended to be helpful for MUAs that show only the final error to their
2836 PH/42 Another patch from the Sieve maintainer.
2838 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2839 Eximstats - Fixed some broken HTML links and added missing column headers
2841 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2842 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2844 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2846 PH/43 Yet another patch from the Sieve maintainer.
2848 PH/44 I found a way to check for a TCP/IP connection going away before sending
2849 the response to the final '.' that terminates a message, but only in the
2850 case where the client has not sent further data following the '.'
2851 (unfortunately, this is allowed). However, in many cases there won't be
2852 any further data because there won't be any more messages to send. A call
2853 to select() can be used: if it shows that the input is "ready", there is
2854 either input waiting, or the socket has been closed. An attempt to read
2855 the next input character can distinguish the two cases. Previously, Exim
2856 would have sent an OK response which the client would never have see.
2857 This could lead to message repetition. This fix should cure that, at
2858 least in a lot of common cases.
2860 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2861 advertised in response to EHLO.
2867 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2868 fixed by 4.65/MH/01 (is this a record?) are fixed:
2870 (i) An empty string was always treated as zero by the numeric comparison
2871 operators. This behaviour has been restored.
2873 (ii) It is documented that the numeric comparison operators always treat
2874 their arguments as decimal numbers. This was broken in that numbers
2875 starting with 0 were being interpreted as octal.
2877 While fixing these problems I realized that there was another issue that
2878 hadn't been noticed. Values of message_size_limit (both the global option
2879 and the transport option) were treated as octal if they started with 0.
2880 The documentation was vague. These values are now always treated as
2881 decimal, and I will make that clear in the documentation.
2887 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2888 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2891 MH/01 Don't check that the operands of numeric comparison operators are
2892 integers when their expansion is in "skipping" mode (fixes bug
2893 introduced by 4.64-PH/07).
2895 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2896 child addresses, Exim now panics and dies. Previously, because the count
2897 is held in a short int, deliveries were likely to be lost. As such a
2898 large number of recipients for a single message is ridiculous
2899 (performance will be very, very poor), I have chosen to impose a limit
2900 rather than extend the field.
2906 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2907 leftover -K file (the existence of which was triggered by #402).
2908 While we were at it, introduced process PID as part of the -K
2909 filename. This should rule out race conditions when creating
2912 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2913 processing considerably. Previous code took too long for large mails,
2914 triggering a timeout which in turn triggers #401.
2916 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2917 in the DK code in transports.c. sendfile() is not really portable,
2918 hence the _LINUX specificness.
2920 TF/01 In the add_headers option to the mail command in an Exim filter,
2921 there was a bug that Exim would claim a syntax error in any
2922 header after the first one which had an odd number of characters
2925 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2926 callout verification, Exim cached a "reject" for the entire domain. This
2927 is correct for most verifications, but it is not correct for a recipient
2928 verification with use_sender or use_postmaster set, because in that case
2929 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2930 case of MAIL FROM:<> rejection from other early rejections (e.g.
2931 rejection of HELO). When verifying a recipient using a non-null MAIL
2932 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2933 Whatever the result of the callout, the value of the domain cache is
2934 left unchanged (for any other kind of callout, getting as far as trying
2935 RCPT means that the domain itself is ok).
2937 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2940 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2941 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2942 session, the daemon ignores ECONNECT errors and logs others; it now
2943 ignores EPIPE as well.
2945 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2946 (quoted-printable decoding).
2948 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2949 later the small subsequent patch to fix an introduced bug.
2951 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2953 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2955 PH/08 An error is now given if message_size_limit is specified negative.
2957 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2958 to be given (somewhat) arbitrary names.
2960 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2963 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2964 miscellaneous code fixes
2966 PH/10 Added the log_reject_target ACL modifier to specify where to log
2969 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2970 hostname. This is wrong, because it relates to the incoming message (and
2971 probably the interface on which it is arriving) and not to the outgoing
2972 callout (which could be using a different interface). This has been
2973 changed to use the value of the helo_data option from the smtp transport
2974 instead - this is what is used when a message is actually being sent. If
2975 there is no remote transport (possible with a router that sets up host
2976 addresses), $smtp_active_hostname is used.
2978 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2979 tweaks were necessary in order to get it to work (see also 21 below):
2980 (a) The code assumed that strncpy() returns a negative number on buffer
2981 overflow, which isn't the case. Replaced with Exim's string_format()
2983 (b) There were several signed/unsigned issues. I just did the minimum
2984 hacking in of casts. There is scope for a larger refactoring.
2985 (c) The code used strcasecmp() which is not a standard C function.
2986 Replaced with Exim's strcmpic() function.
2987 (d) The code set only $1; it now sets $auth1 as well.
2988 (e) A simple test gave the error "authentication client didn't specify
2989 service in request". It would seem that Dovecot has changed its
2990 interface. Fortunately there's a specification; I followed it and
2991 changed what the client sends and it appears to be working now.
2993 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2996 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2997 address A is aliased to B and C, where B exists and C does not. Without
2998 -v the output is "A verified" because verification stops after a
2999 successful redirection if more than one address is generated. However,
3000 with -v the child addresses are also verified. Exim was outputting "A
3001 failed to verify" and then showing the successful verification for C,
3002 with its parentage. It now outputs "B failed to verify", showing B's
3003 parentage before showing the successful verification of C.
3005 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3006 look up a TXT record in a specific list after matching in a combined
3009 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3010 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3011 they consult the DNS. I had assumed they would set it the way they
3012 wanted; and indeed my experiments on Linux seem to show that in some
3013 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3014 To be on the safe side, however, I have now made the interface to
3015 host_find_byname() similar to host_find_bydns(), with an argument
3016 containing the DNS resolver options. The host_find_byname() function now
3017 sets these options at its start, just as host_find_bydns() does. The smtp
3018 transport options dns_qualify_single and dns_search_parents are passed to
3019 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3020 of host_find_byname() use the default settings of RES_DEFNAMES
3021 (qualify_single) but not RES_DNSRCH (search_parents).
3023 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3024 spool_read_header() do less string testing, by means of a preliminary
3025 switch on the second character of optional "-foo" lines. (This is
3026 overdue, caused by the large number of possibilities that now exist.
3027 Originally there were few.) While I was there, I also converted the
3028 str(n)cmp tests so they don't re-test the leading "-" and the first
3029 character, in the hope this might squeeze out yet more improvement.
3031 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3032 flag allowing group syntax was set by the header_syntax check but not
3033 turned off, possible causing trouble later; (2) The flag was not being
3034 set at all for the header_verify test, causing "group"-style headers to
3035 be rejected. I have now set it in this case, and also caused header_
3036 verify to ignore an empty address taken from a group. While doing this, I
3037 came across some other cases where the code for allowing group syntax
3038 while scanning a header line wasn't quite right (mostly, not resetting
3039 the flag correctly in the right place). These bugs could have caused
3040 trouble for malformed header lines. I hope it is now all correct.
3042 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3043 with the "reply" argument non-NULL. The code, however (which originally
3044 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3045 but it didn't always do it. This confused somebody who was copying the
3046 code for some other use. I have removed all the tests.
3048 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3049 feature that was used to support insecure browsers during the U.S. crypto
3050 embargo. It requires special client support, and Exim is probably the
3051 only MTA that supported it -- and would never use it because real RSA is
3052 always available. This code has been removed, because it had the bad
3053 effect of slowing Exim down by computing (never used) parameters for the
3054 RSA_EXPORT functionality.
3056 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3057 authenticator to fail if there's a tab character in the incoming data
3058 (there should never be unless someone is messing about, as it's supposed
3059 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3060 if the connection is using TLS or if the remote IP is the same as the
3061 local IP, and the "valid-client-cert option" if a client certificate has
3064 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3065 authenticators. This can be used for authorization after authentication
3066 succeeds. (In the case of plaintext, it servers for both authentication
3069 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3070 if any retry times were supplied.
3072 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3073 connection, where there is no client IP address to check. In this
3074 situation, the verify now always succeeds.
3076 PH/25 Applied John Jetmore's -Mset patch.
3078 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3080 PH/27 In a string expansion for a processed (not raw) header when multiple
3081 headers of the same name were present, leading whitespace was being
3082 removed from all of them, but trailing whitespace was being removed only
3083 from the last one. Now trailing whitespace is removed from each header
3084 before concatenation. Completely empty headers in a concatenation (as
3085 before) are ignored.
3087 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3088 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3090 PH/29 [Removed. This was a change that I later backed out, and forgot to
3091 correct the ChangeLog entry (that I had efficiently created) before
3092 committing the later change.]
3094 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3095 address errors (4xx response to RCPT) over the same connection as other
3096 messages routed to the same hosts. Such deliveries are always "forced",
3097 so retry times are not inspected. This resulted in far too many retries
3098 for the affected addresses. The effect occurred only when there were more
3099 hosts than the hosts_max_try setting in the smtp transport when it had
3100 the 4xx errors. Those hosts that it had tried were not added to the list
3101 of hosts for which the message was waiting, so if all were tried, there
3102 was no problem. Two fixes have been applied:
3104 (i) If there are any address or message errors in an SMTP delivery, none
3105 of the hosts (tried or untried) are now added to the list of hosts
3106 for which the message is waiting, so the message should not be a
3107 candidate for sending over the same connection that was used for a
3108 successful delivery of some other message. This seems entirely
3109 reasonable: after all the message is NOT "waiting for some host".
3110 This is so "obvious" that I'm not sure why it wasn't done
3111 previously. Hope I haven't missed anything, but it can't do any
3112 harm, as the worst effect is to miss an optimization.
3114 (ii) If, despite (i), such a delivery is accidentally attempted, the
3115 routing retry time is respected, so at least it doesn't keep
3116 hammering the server.
3118 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3119 in ${readsocket because some servers need this prod.
3121 PH/32 Added some extra debug output when updating a wait-xxx database.
3123 PH/33 The hint "could be header name not terminated by colon", which has been
3124 given for certain expansion errors for a long time, was not being given
3125 for the ${if def:h_colon_omitted{... case.
3127 PH/34 The spec says: "With one important exception, whenever a domain list is
3128 being scanned, $domain contains the subject domain." There was at least
3129 one case where this was not true.
3131 PH/35 The error "getsockname() failed: connection reset by peer" was being
3132 written to the panic log as well as the main log, but it isn't really
3133 panic-worthy as it just means the connection died rather early on. I have
3134 removed the panic log writing for the ECONNRESET error when getsockname()
3137 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3138 runs only) independently of the message's sender address. This meant
3139 that, if the 4xx error was in fact related to the sender, a different
3140 message to the same recipient with a different sender could confuse
3141 things. In particular, this can happen when sending to a greylisting
3142 server, but other circumstances could also provoke similar problems.
3143 I have changed the default so that the retry time for these errors is now
3144 based a combination of the sender and recipient addresses. This change
3145 can be overridden by setting address_retry_include_sender=false in the
3148 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3149 remote server are returned as part of bounce messages. This was not
3150 happening for LMTP over a pipe (the lmtp transport), but now it is the
3151 same for both kinds of LMTP.
3153 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3154 in header lines that were in fact CNAMEs. This is no longer the case.
3156 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3157 and queue runs started by the daemon processed all messages. This has
3158 been fixed so that -R and -S can now usefully be given with -q<time>.
3160 PH/40 Import PCRE release 6.7 (fixes some bugs).
3162 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3164 PH/42 Give an error if -q is specified more than once.
3166 PH/43 Renamed the variables $interface_address and $interface_port as
3167 $received_ip_address and $received_port, to make it clear that these
3168 values apply to message reception, and not to the outgoing interface when
3169 a message is delivered. (The old names remain recognized, of course.)
3171 PH/44 There was no timeout on the connect() call when using a Unix domain
3172 socket in the ${readsocket expansion. There now is.
3174 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3175 be meaningful with "accept".
3177 SC/01 Eximstats V1.43
3178 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3180 SC/02 Eximstats V1.44
3181 Use a glob alias rather than an array ref in the generated
3182 parser. This improves both readability and performance.
3184 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3185 Collect SpamAssassin and rejection statistics.
3186 Don't display local sender or destination tables unless
3187 there is data to show.
3188 Added average volumes into the top table text output.
3190 SC/04 Eximstats V1.46
3191 Collect data on the number of addresses (recipients)
3192 as well as the number of messages.
3194 SC/05 Eximstats V1.47
3195 Added 'Message too big' to the list of mail rejection
3196 reasons (thanks to Marco Gaiarin).
3198 SC/06 Eximstats V1.48
3199 Mainlog lines which have GMT offsets and are too short to
3200 have a flag are now skipped.
3202 SC/07 Eximstats V1.49 (Alain Williams)
3203 Added the -emptyok flag.
3205 SC/08 Eximstats V1.50
3206 Fixes for obtaining the IP address from reject messages.
3208 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3209 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3210 whitespace changes from 4.64-PH/27
3212 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3215 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3216 are found, allow negative numbers in numeric criteria)
3218 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3220 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3221 to match changes made in 4.64-PH/43
3223 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3225 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3226 "domain" is actually an IP address, and force a failure. This locks out
3227 those revolvers/nameservers that support "A-for-A" lookups, in
3228 contravention of the specifications.
3230 PH/48 When a host name was looked up from an IP address, and the subsequent
3231 forward lookup of the name timed out, the host name was left in
3232 $sender_host_name, contrary to the specification.
3234 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3235 restricted to single-key lookups, Exim was not diagnosing an error if
3236 * or *@ was used with a query-style lookup.
3238 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3240 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3241 long ago, but noone interested enough thought of it. Let's just say that
3242 the "1.1" means that there are some new functions that weren't there at
3243 some point in the past.
3245 PH/51 Error processing for expansion failure of helo_data from an smtp
3246 transport during callout processing was broken.
3248 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3249 tested/used via the -bh/-bhc/-bs options.
3251 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3252 bug, fixed in subsequent PCRE releases).
3254 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3255 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3257 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3263 SC/01 Use a glob alias rather than an array ref in eximstats generated
3264 parser. This improves both readability and performance.
3266 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3267 Don't display local sender or destination tables in eximstats unless
3268 there is data to show.
3269 Added average volumes into the eximstats top table text output.
3271 SC/03 Collect data on the number of addresses (recipients) as well
3272 as the number of messages in eximstats.
3274 TF/01 Correct an error in the documentation for the redirect router. Exim
3275 does (usually) call initgroups() when daemonizing.
3277 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3278 with consistent privilege compared to when running as a daemon.
3280 TF/03 Note in the spec that $authenticated_id is not set for local
3281 submissions from trusted users.
3283 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3284 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3286 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3287 by adding some example configuration directives to the default
3288 configuration file. A little bit of work is required to uncomment the
3289 directives and define how usernames and passwords are checked, but
3290 there is now a framework to start from.
3292 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3293 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3294 without this. I don't know how relevant this is to other LDAP libraries.
3296 PH/02 Add the verb name to the "unknown ACL verb" error.
3298 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3300 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3302 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3303 directory not expanded when it should be if an expanded home directory
3304 was set for the address (which is overridden by the transport).
3306 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3309 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3310 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3311 because it is too late at that time, and has no effect.
3313 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3314 security issue with \' (bugzilla #107). I could not use the
3315 PQescapeStringConn() function, because it needs a PGconn value as one of
3318 PH/08 When testing addresses using -bt, indicate those final addresses that
3319 are duplicates that would not cause an additional delivery. At least one
3320 person was confused, thinking that -bt output corresponded to deliveries.
3321 (Suppressing duplicates isn't a good idea as you lose the information
3322 about possibly different redirections that led to the duplicates.)
3324 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3325 systems where poll() doesn't work, in particular OS X.
3327 PH/10 Added more information to debugging output for retry time not reached.
3329 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3330 operations in malware.c.
3332 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3335 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3336 syslog_duplication set false, log lines that would normally be written
3337 both the the main log and to the reject log were not written to syslog at
3340 PH/14 In the default configuration, change the use of "message" in ACL warn
3341 statements to "add_header".
3343 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3344 not followed by a command (e.g. "seen endif").
3346 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3347 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3350 PH/17 Added extra conditions to the default value of delay_warning_condition
3354 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3355 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3356 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3359 The Auto-Submitted: and various List- headers are standardised, whereas I
3360 don't think Precedence: ever was.
3362 PH/18 Refactored debugging code in route_finduser() to show more information,
3363 in particular, the error code if getpwnam() issues one.
3365 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3366 This is apparently needed in addition to the PH/07 change above to avoid
3367 any possible encoding problems.
3369 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3370 but not after initializing Perl.
3372 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3373 output them only if debugging. By default they are written stderr,
3374 apparently, which is not desirable.
3376 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3379 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3382 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3384 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3385 authenticated or an ident call has been made. Suppress the default
3386 values for $authenticated_id and $authenticated_sender (but permit -oMai
3387 and -oMas) when testing with -bh.
3389 PH/24 Re-jigged the order of the tests in the default configuration so that the
3390 tests for valid domains and recipients precede the DNS black list and CSA
3391 tests, on the grounds that those ones are more expensive.
3393 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3394 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3395 This bug exists in every version of Exim that I still have, right back to
3398 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3399 However, an attempt to turn on case-sensitivity in a regex key by
3400 including (?-i) didn't work because the subject string was already
3401 lowercased, and the effects were non-intuitive. It turns out that a
3402 one-line patch can be used to allow (?-i) to work as expected.
3408 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3409 other effects) broke the use of negated acl sub-conditions.
3411 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3414 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3415 "Deny" causes Exim to reject the incoming connection with a 554 error.
3416 Unfortunately, if there is a major crisis, such as a disk failure,
3417 tcp-wrappers gives "deny", whereas what one would like would be some
3418 kind of temporary error. A kludge has been added to help with this.
3419 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3420 554 error is used if errno is still zero or contains ENOENT (which occurs
3421 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3424 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3426 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3427 errors. Otherwise a message that provokes a temporary error (when other
3428 messages do not) can cause a whole host to time out.
3430 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3431 addresses were routed directly to files or pipes from a redirect router.
3432 File deliveries just didn't batch; pipe deliveries might have suffered
3435 PH/06 A failure to get a lock for a hints database would erroneously always say
3436 "Failed to get write lock", even when it was really a read lock.
3438 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3439 of 0600. This has been changed to use the value of the lockfile_mode
3440 option (which defaults to 0600).
3442 PH/08 Applied small patch from the Sieve maintainer.
3444 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3445 folder from quota calculations, a direct delivery into this folder messed
3446 up the contents of the maildirsize file. This was because the regex was
3447 used only to exclude .Trash (or whatever) when the size of the mailbox
3448 was calculated. There was no check that a delivery was happening into an
3449 excluded directory. This bug has been fixed by ignoring all quota
3450 processing for deliveries into excluded directories.
3452 PH/10 Added the maildirfolder_create_regex option to appendfile.
3458 PH/01 The code for finding all the local interface addresses on a FreeBSD
3459 system running IPv6 was broken. This may well have applied to all BSD
3460 systems, as well as to others that have similar system calls. The broken
3461 code found IPv4 interfaces correctly, but gave incorrect values for the
3462 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3463 that it would not match correctly against @[] and not recognize the IPv6
3466 PH/02 The ipliteral router was not recognizing addresses of the form user@
3467 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3469 PH/03 Added disable_ipv6.
3471 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3472 decoded form, because it is most often used to construct To: headers
3473 lines in autoreplies, and the decoded form may well be syntactically
3474 invalid. However, $reply_address has leading white space removed, and all
3475 newlines turned into spaces so that the autoreply transport does not
3478 PH/05 If group was specified without a user on a router, and no group or user
3479 was specified on a transport, the group from the router was ignored.
3481 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3482 for visible compile-time settings that can be used to change these
3483 numbers, for those that want even more. Backwards compatibility with old
3484 spool files has been maintained. However, going back to a previous Exim
3485 release will lost any variables that are in spool files.
3487 PH/07 Two small changes when running in the test harness: increase delay when
3488 passing a TCP/IP connection to a new process, in case the original
3489 process has to generate a bounce, and remove special handling of
3490 127.0.0.2 (sic), which is no longer necessary.
3492 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3493 be the same on different OS.
3495 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3498 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3499 whether --show-vars was specified or not
3501 JJ/02 exipick: Added support for new ACL variable spool format introduced
3504 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3505 syntactically invalid From: or Reply-to: line, and a filter used this to
3506 generate an autoreply, and therefore failed to obtain an address for the
3507 autoreply, Exim could try to deliver to a non-existent relative file
3508 name, causing unrelated and misleading errors. What now happens is that
3509 it logs this as a hard delivery error, but does not attempt to create a
3512 PH/11 The exinext utility has a -C option for testing purposes, but although
3513 the given file was scanned by exinext itself; it wasn't being passed on
3514 when Exim was called.
3516 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3517 an end-of-file indication when reading a command response.
3519 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3520 compiled. In many other places in Exim, IPv6 addresses are always
3521 recognized, so I have changed this. It also means that IPv4 domain
3522 literals of the form [IPV4:n.n.n.n] are now always recognized.
3524 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3525 used if the router is not running as root, for example, when verifying at
3526 ACL time, or when using -bh. The debugging output from this situation was
3527 non-existent - all you got was a failure to exec. I have made two
3530 (a) Failures to set uid/gid, the current directory, or a process leader
3531 in a subprocess such as that created by queryprogram now generate
3532 suitable debugging output when -d is set.
3534 (b) The queryprogram router detects when it is not running as root,
3535 outputs suitable debugging information if -d is set, and then runs
3536 the subprocess without attempting to change uid/gid.
3538 PH/15 Minor change to Makefile for building test_host (undocumented testing
3541 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3542 additional section of a DNS packet that returns MX or SRV records.
3543 Instead, it always explicitly searches for A/AAAA records. This avoids
3544 major problems that occur when a DNS server includes only records of one
3545 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3546 fixed another bug: if SRV records were looked up and the corresponding
3547 address records were *not* found in the additional section, the port
3548 values from the SRV records were lost.
3550 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3551 using the correct key (the original address) when searching the retry
3552 rules in order to find which one to use for generating the retry hint.
3554 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3555 adding the default one. Similarly, if it contains a Reply-To: header, the
3556 errors_reply_to option, if set, is not used.
3558 PH/19 When calculating a retry time, Exim used to measure the "time since
3559 failure" by looking at the "first failed" field in the retry record. Now
3560 it does not use this if it is later than than the arrival time of the
3561 message. Instead it uses the arrival time. This makes for better
3562 behaviour in cases where some deliveries succeed, thus re-setting the
3563 "first failed" field. An example is a quota failure for a huge message
3564 when small messages continue to be delivered. Without this change, the
3565 "time since failure" will always be short, possible causing more frequent
3566 delivery attempts for the huge message than are intended.
3567 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3569 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3570 $1, $2, $3) because the numerical variables can be reset during some
3571 expansion items (e.g. "match"), thereby losing the authentication data.
3573 PH/21 Make -bV show the size of off_t variables so that the test suite can
3574 decide whether to run tests for quotas > 2G.
3576 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3577 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3578 filecount value is greater than 2G or if a quota value is greater than 2G
3579 on a system where the size of off_t is not greater than 4, a panic error
3582 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3583 never match. The debug and -bh output now contains an explicit error
3584 message indicating a malformed IPv4 address or mask.
3586 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3587 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3588 PH/23 above applies.
3590 PH/25 Do not write to syslog when running in the test harness. The only
3591 occasion when this arises is a failure to open the main or panic logs
3592 (for which there is an explicit test).
3594 PH/26 Added the /no_tell option to "control=freeze".
3596 PH/27 If a host name lookup failed very early in a connection, for example, if
3597 the IP address matched host_lookup and the reverse lookup yielded a name
3598 that did not have a forward lookup, an error message of the form "no IP
3599 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3600 could be logged. Now it outputs the IP address instead of "NULL".
3602 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3603 allows the sender and the authenticated sender to be set when
3604 submitting a message from within Exim. Since child_open_exim() is
3605 documented for local_scan(), the new function should be too.
3607 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3608 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3609 results in an empty string is now treated as unset.
3611 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3613 PH/31 Added sender_verify_fail logging option.
3615 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3616 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3617 tidied the source and removed it altogether.
3619 PH/33 When a queue run was abandoned because the load average was too high, a
3620 log line was always written; now it is written only if the queue_run log
3621 selector is set. In addition, the log line for abandonment now contains
3622 information about the queue run such as the pid. This is always present
3623 in "start" and "stop" lines but was omitted from the "abandon" line.
3625 PH/34 Omit spaces between a header name and the colon in the error message that
3626 is given when verify = headers_syntax fails (if there are lots of them,
3627 the message gets confusing).
3629 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3630 names, as there are now some PTR records that contain slashes. This check
3631 is only to protect against broken name servers that fall over on strange
3632 characters, so the fact that it applies to all lookups doesn't matter.
3634 PH/36 Now that the new test suite is complete, we can remove some of the
3635 special code in Exim that was needed for the old test suite. For example,
3636 sorting DNS records because real resolvers return them in an arbitrary
3637 order. The new test suite's fake resolver always returns records in the
3640 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3641 bounces) except when queue_only is set, to avoid logging races between
3642 the different processes.
3644 PH/38 Panic-die if .include specifies a non-absolute path.
3646 PH/39 A tweak to the "H" retry rule from its user.
3648 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3649 a label. They prevented compilation on older perls.
3651 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3652 a warning to be raised on newish perls.
3654 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3655 on queue. Changes to match documented behaviour of showing count of
3656 messages matching specified criteria.
3658 PH/40 Changed the default ident timeout from 30s to 5s.
3660 PH/41 Added support for the use of login_cap features, on those BSD systems
3661 that have them, for controlling the resources used by pipe deliveries.
3663 PH/42 The content-scanning code uses fopen() to create files in which to put
3664 message data. Previously it was not paying any attention to the mode of
3665 the files. Exim runs with umask(0) because the rest of the code creates
3666 files with open(), and sets the required mode explicitly. Thus, these
3667 files were ending up world-writeable. This was not a big issue, because,
3668 being within the spool directory, they were not world-accessible. I have
3669 created a function called modefopen, which takes an additional mode
3670 argument. It sets umask(777), creates the file, chmods it to the required
3671 mode, then resets the umask. All the relevant calls to fopen() in the
3672 content scanning code have been changed to use this function.
3674 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3675 to 24 hours. This avoids potential overflow problems when processing G
3676 and H retry rules. I suspect nobody ever tinkers with this value.
3678 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3680 PH/45 When the plaintext authenticator is running as a client, the server's
3681 challenges are checked to ensure they are valid base64 strings. By
3682 default, the authentication attempt is cancelled if an invalid string is
3683 received. Setting client_ignore_invalid_base64 true ignores these errors.
3684 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3685 they are received. Thus, the responses can be made to depend on the
3686 challenges. If an invalid string is ignored, an empty string is placed in
3689 PH/46 Messages that are created by the autoreply transport now contains a
3690 References: header, in accordance with RFCs 2822 and 3834.
3692 PH/47 Added authenticated_sender_force to the smtp transport.
3694 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3696 PH/49 Installed latest patch from the Sieve maintainer.
3698 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3699 was also set, the appendfile transport was unnecessarily scanning a
3700 directory of message files (e.g. for maildir delivery) to find the count
3701 of files (along with the size), even though it did not need this
3702 information. It now does the scan only if it needs to find either the
3703 size of the count of files.
3705 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3707 PH/52 Two bugs concerned with error handling when the smtp transport is
3710 (i) Exim was not creating retry information for temporary errors given
3711 for individual recipients after the DATA command when the smtp transport
3712 was used in LMTP mode. This meant that they could be retried too
3713 frequently, and not timed out correctly.
3715 (ii) Exim was setting the flag that allows error details to be returned
3716 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3717 recipients that were returned after the DATA command.
3719 PH/53 This is related to PH/52, but is more general: for any failing address,
3720 when detailed error information was permitted to be returned to the
3721 sender, but the error was temporary, then after the final timeout, only
3722 "retry timeout exceeded" was returned. Now it returns the full error as
3723 well as "retry timeout exceeded".
3725 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3726 do this, and (what is worse) MTAs that accept it.
3728 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3729 will now be deprecated.
3731 PH/56 New os.c-cygwin from the Cygwin maintainer.
3733 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3734 formats (previously only available in exim formats via -bpr, -bpru,
3735 and -bpra. Now also available in native and exiqgrep formats)
3737 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3738 with very large, slow to parse queues
3740 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3742 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3744 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3745 responses to authentication challenges, though it was showing the
3746 challenges; (ii) I've removed the CR characters from the debug output for
3749 PH/58 Allow for the insertion of a newline as well as a space when a string
3750 is turned into more than one encoded-word during RFC 2047 encoding. The
3751 Sieve code now uses this.
3753 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3754 data_4xx, lost_connection, tls_required.
3756 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3757 message was being sent as an SMTP response.
3759 PH/61 Add -l and -k options to exicyclog.
3761 PH/62 When verifying, if an address was redirected to one new address, so that
3762 verification continued, and the new address failed or deferred after
3763 having set something in $address_data, the value of $address_data was not
3764 passed back to the ACL. This was different to the case when no
3765 redirection occurred. The value is now passed back in both cases.
3767 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3768 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3769 use login_cap.h, so on its own it isn't the distinguishing feature. The
3770 new name refers directly to the setclassresources() function.
3772 PH/65 Added configuration files for NetBSD3.
3774 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3776 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3777 is preferred over IPv4.
3779 PH/68 The bounce_return_message and bounce_return_body options were not being
3780 honoured for bounces generated during the reception of non-SMTP messages.
3781 In particular, this applied to messages rejected by the ACL. This bug has
3782 been fixed. However, if bounce_return_message is true and bounce_return_
3783 body is false, the headers that are returned for a non-SMTP message
3784 include only those that have been read before the error was detected.
3785 (In the case of an ACL rejection, they have all been read.)
3787 PH/69 The HTML version of the specification is now built in a directory called
3788 spec_html instead of spec.html, because the latter looks like a path with
3789 a MIME-type, and this confuses some software.
3791 PH/70 Catch two compiler warnings in sieve.c.
3793 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3794 function verify_get_ident() calls ip_connect() to connect a socket, but
3795 if the "connect()" function timed out, ip_connect() used to close the
3796 socket. However, verify_get_ident() also closes the socket later, and in
3797 between Exim writes to the log, which may get opened at this point. When
3798 the socket was closed in ip_connect(), the log could get the same file
3799 descriptor number as the socket. This naturally causes chaos. The fix is
3800 not to close the socket in ip_connect(); the socket should be closed by
3801 the function that creates it. There was only one place in the code where
3802 this was missing, in the iplookup router, which I don't think anybody now
3803 uses, but I've fixed it anyway.
3805 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3806 well as to direct DNS lookups. Otherwise the handling of names in host
3807 lists is inconsistent and therefore confusing.
3813 PH/01 Two changes to the default runtime configuration:
3815 (1) Move the checks for relay_from_hosts and authenticated clients from
3816 after to before the (commented out) DNS black list checks.
3818 (2) Add control=submission to the relay_from_hosts and authenticated
3819 clients checks, on the grounds that messages accepted by these
3820 statements are most likely to be submissions.
3822 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3824 (1) Generate an error if the third argument for the ${prvs expansion is
3827 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3830 (3) Reset the variables that are obtained from the first argument of
3831 ${prvscheck and used in the second argument before leaving the code,
3832 because their memory is reclaimed, so using them afterwards may do
3835 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3836 one (it's much easier than Tom thought :-).
3838 (5) Because of (4), we can now allow for the use of $prvscheck_result
3839 inside the third argument.
3841 PH/03 For some reason, the default setting of PATH when running a command from
3842 a pipe transport was just "/usr/bin". I have changed it to
3845 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3846 anything to be listed in the output from -bV.
3848 PH/05 When a filter generated an autoreply, the entire To: header line was
3849 quoted in the delivery log line, like this:
3851 => >A.N.Other <ano@some.domain> <original@ddress> ...
3853 This has been changed so that it extracts the operative address. There
3854 may be more than one such address. If so, they are comma-separated, like
3857 => >ano@some.domain,ona@other.domain <original@ddress> ...
3859 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3860 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3861 not being looked up in the rDNS to get a host name, Exim was showing the
3862 IP address twice in Received: lines, even though the IP addresses were
3863 identical. For example:
3865 Received: from [1.2.3.4] (helo=[1.2.3.4])
3867 However, if the real host name was known, it was omitting the HELO data
3868 if it matched the actual IP address. This has been tidied up so that it
3869 doesn't show the same IP address twice.
3871 PH/07 When both +timestamp and +memory debugging was on, the value given by
3872 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3873 called by the debug printing, thereby overwriting the timestamp buffer.
3874 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3876 PH/08 When the original message was included in an autoreply transport, it
3877 always said "this is a copy of the message, including all the headers",
3878 even if body_only or headers_only was set. It now gives an appropriate
3881 PH/09 Applied a patch from the Sieve maintainer which:
3883 o fixes some comments
3884 o adds the (disabled) notify extension core
3885 o adds some debug output for the result of if/elsif tests
3886 o points to the current vacation draft in the documentation
3887 and documents the missing references header update
3891 o fixes a bug in processing the envelope test (when testing
3892 multiple envelope elements, the last element determined the
3895 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3896 Electronic Mail") by including:
3898 Auto-submitted: auto-generated
3900 in the messages that it generates (bounce messages and others, such as
3901 warnings). In the case of bounce messages for non-SMTP messages, there was
3902 also a typo: it was using "Auto_submitted" (underscore instead of
3903 hyphen). Since every message generated by Exim is necessarily in response
3904 to another message, thes have all been changed to:
3906 Auto-Submitted: auto-replied
3908 in accordance with these statements in the RFC:
3910 The auto-replied keyword:
3912 - SHOULD be used on messages sent in direct response to another
3913 message by an automatic process,
3915 - MUST NOT be used on manually-generated messages,
3917 - MAY be used on Delivery Status Notifications (DSNs) and Message
3918 Disposition Notifications (MDNs),
3920 - MUST NOT be used on messages generated by automatic or periodic
3921 processes, except for messages which are automatic responses to
3924 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3925 to the default Received: header definition.
3927 PH/12 Added log selector acl_warn_skipped (default on).
3929 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3930 variables because (a) they are in the wrong storage pool and (b) even if
3931 they were copied, it wouldn't work properly because of the caching.
3933 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3934 checking when decoding. Apparently there are clients that generate
3935 overlong encoded strings. Why am I not surprised?
3937 PH/15 If the first argument of "${if match_address" was not empty, but did not
3938 contain an "@" character, Exim crashed. Now it writes a panic log message
3939 and treats the condition as false.
3941 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3943 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3944 extension "envelope-auth". The code is finished and in agreement with
3945 other implementations, but there is no documentation so far and in fact,
3946 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3947 not changing the active code.
3949 Print executed "if" and "elsif" statements when debugging is used. This
3950 helps a great deal to understand what a filter does.
3952 Document more things not specified clearly in RFC3028. I had all this
3953 sorted out, when out of a sudden new issues came to my mind. Oops."
3955 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3958 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3959 canonical form (as documented). However, after a host name lookup from
3960 the IP address, check_host() was doing a simple string comparison with
3961 addresses acquired from the DNS when checking that the found name did
3962 have the original IP as one of its addresses. Since any found IPv6
3963 addresses are likely to be in abbreviated form, the comparison could
3964 fail. Luckily, there already exists a function for doing the comparison
3965 by converting both addresses to binary, so now that is used instead of
3966 the text comparison.
3968 PH/20 There was another similar case to PH/19, when a complete host name was
3969 given in a host list; looking up its IP address could give an abbreviated
3970 form, whereas the current host's name might or might not be abbreviated.
3971 The same fix has been applied.
3977 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3978 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3981 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3982 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3984 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3986 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3987 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3988 Also, for "Auto-Submitted", treat anything other than "no" as
3989 non-personal, in accordance with RFC 3834. (Previously it treated
3990 anything starting "auto-" as non-personal.)
3992 TF/01 The control=submission/name=... option had a problem with syntax
3993 errors if the name included a slash character. The /name= option
3994 now slurps the rest of the string, so it can include any characters
3995 but it must come last in the list of options (after /sender_retain
3998 PH/05 Some modifications to the interface to the fake nameserver for the new
4006 TK/01 Added the "success_on_redirect" address verification option. See
4007 NewStuff for rationale and an example.
4009 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4011 PH/02 Patch to exigrep to allow it to work on syslog lines.
4013 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4014 fread() to skip over the body file's header line, because in Cygwin the
4015 header line is locked and is inaccessible.
4017 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4018 co-exist for some time) to make it clear that it is the Exim ID that is
4019 referenced, not the Message-ID: header line.
4021 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4022 string_format() function, because snprintf() does not exist on all
4025 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4026 ${stat: expansion item.
4028 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4029 protocol synchronization error", to keep the pedants happy.
4031 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4032 well as for IRIX systems, when gcc is being used. See the host.c source
4035 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4037 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4040 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4041 transport and to the smtp transport in LMTP mode.
4043 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4045 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4046 run a filter in a subprocess. This could lead to confusion in subsequent
4047 lookups in the parent process. There should also be a search_tidyup() at
4048 the end of the subprocess.
4050 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4051 only if the host matched helo_try_verify_hosts, which caused the
4052 verification to occur when the EHLO/HELO command was issued. The ACL just
4053 tested the remembered result. Now, if a previous verification attempt has
4054 not happened, "verify = helo" does it there and then.
4056 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4058 TK/03 Fix log output including CR from clamd.
4060 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4061 exist provoked a memory error which could cause a segfault.
4063 PH/15 Installed PCRE 6.2
4065 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4067 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4068 of the problem. Specifically, suggested +O2 rather than +O1 for the
4071 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4073 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4074 record, the port was not correctly logged when the outgoing_port log
4075 selector was set (it logged the transort's default port).
4077 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4078 fallback_hosts, and "hosts" in the smtp transport.
4080 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4081 host errors such as "Connection refused".
4083 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4084 authentication with radiusclient 0.4.9:
4086 - Error returned from rc_read_config was caught wrongly
4087 - Username/password not passed on to radius server due to wrong length.
4089 The presumption is that some radiusclient API changes for 4.51/PH/17
4090 were not taken care of correctly. The code is still untested by me (my
4091 Linux distribution still has 0.3.2 of radiusclient), but it was
4092 contributed by a Radius user.
4094 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4095 expanding the "port" option of the smtp transport.
4097 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4098 while reading a MIME header. Thanks to Tom Hughes for a patch.
4100 PH/24 Include config.h inside local_scan.h so that configuration settings are
4103 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4104 that in an ACL for RCPT (for example), you can examine exactly what was
4107 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4108 commands, but it was not correctly comparing the address with the actual
4109 client host address. Thus, it would show the EHLO address in Received:
4110 header lines when this was not necessary.
4112 PH/27 Added the % operator to ${eval:}.
4114 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4115 it should be ignoring failures (because with -C, for example, it has lost
4116 privilege). It wasn't ignoring creation failures other than "already
4119 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4122 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4123 because an input line was too long, either on its own, or by virtue of
4124 too many continuations, the temporary file was not being removed, and the
4125 return code was incorrect.
4127 PH/31 Missing "BOOL" in function definition in filtertest.c.
4129 PH/32 Applied Sieve patches from the maintainer.
4131 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4133 PH/33 Added "verify = not_blind".
4135 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4136 Local/Makefile (with some defaults set). These are used in built scripts
4137 such as exicyclog, but they have never been used in the exim_install
4138 script (though there are many overriding facilities there). I have
4139 arranged that the exim_install script now takes note of these two
4142 PH/35 Installed configuration files for Dragonfly.
4144 PH/36 When a locally submitted message by a trusted user did not contain a
4145 From: header, and the sender address was obtained from -f or from an SMTP
4146 MAIL command, and the trusted user did not use -F to supply a sender
4147 name, $originator_name was incorrectly used when constructing a From:
4148 header. Furthermore, $originator_name was used for submission mode
4149 messages from external hosts without From: headers in a similar way,
4150 which is clearly wrong.
4152 PH/37 Added control=suppress_local_fixups.
4154 PH/38 When log_selector = +received_sender was set, and the addition of the
4155 sender made the log line's construction buffer exactly full, or one byte
4156 less than full, an overflow happened when the terminating "\n" was
4159 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4160 when the result of a list match is failure because a DNS lookup failed.
4162 PH/40 RM_COMMAND is now used in the building process.
4164 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4165 the "build-* directories that it finds.
4167 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4168 address in a domain literal was a prefix of an interface address.
4170 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4171 when verifying a sender address, unless rewrite_headers is false.
4173 PH/44 Wrote a long comment about why errors_to addresses are verified as
4174 recipients, not senders.
4176 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4177 the ratelimit ACL was added.
4179 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4181 PH/46 Added extra information about PostgreSQL errors to the error string.
4183 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4184 avoiding the need to install special zones in a real server. This is
4185 backwards compatible; if it can't find the fake resolver, it drops back.
4186 Thus, both old and new test suites can be run.
4188 TF/02 Added util/ratelimit.pl
4190 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4191 clock is set back in time.
4193 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4194 Candler <B.Candler@pobox.com>.
4196 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4197 OK for addresses that are the result of redirections.
4199 PH/48 A number of further additions for the benefit of the new test suite,
4200 including a fake gethostbyname() that interfaces to the fake DNS resolver
4203 TF/06 The fix for widen_domains has also been applied to qualify_single and
4204 search_parents which are the other dnslookup options that can cause
4207 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4210 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4212 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4213 service" if it encountered an MX record with an empty target hostname.
4214 The message is now "an MX or SRV record indicated no SMTP service".
4216 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4217 if the DNS of the sending site is misconfigured. This is quite a
4218 common situation. This change restores the behaviour of treating a
4219 helo verification defer as a failure.
4221 PH/51 If self=fail was set on a router, the bounce message did not include the
4222 actual error message.
4228 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4230 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4231 command itself ended in error, the underlying message about the transport
4232 filter timeout was being overwritten with the pipe command error. Now the
4233 underlying error message should be appended to the second error message.
4235 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4237 PH/02 Reduce the amount of output that "make" produces by default. Full output
4238 can still be requested.
4240 PH/03 The warning log line about a condition test deferring for a "warn" verb
4241 was being output only once per connection, rather than after each
4242 occurrence (because it was using the same function as for successful
4243 "warn" verbs). This seems wrong, so I have changed it.
4245 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4246 it should not have, which might have caused a crash in the right
4247 circumstances, but probably never did.
4249 PH/04 Installed a modified version of Tony Finch's patch to make submission
4250 mode fix the return path as well as the Sender: header line, and to
4251 add a /name= option so that you can make the user's friendly name appear
4254 TF/03 Added the control = fakedefer ACL modifier.
4256 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4257 Mark Lowes for thorough testing.
4259 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4261 TK/03 Merged latest SRS patch from Miles Wilton.
4263 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4264 with the definition in sysexits.h (which is #included earlier).
4265 Fortunately, Exim does not actually use EX_OK. The code used to try to
4266 preserve the sysexits.h value, by assuming that macro definitions were
4267 scanned for macro replacements. I have been disabused of this notion,
4268 so now the code just undefines EX_OK before #including unistd.h.
4270 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4271 in the smtp transport. When a block could not be written in a single
4272 write() function, the timeout was being re-applied to each part-write.
4273 This seems wrong - if the receiver was accepting one byte at a time it
4274 would take for ever. The timeout is now adjusted when this happens. It
4275 doesn't have to be particularly precise.
4277 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4278 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4280 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4281 without a domain if the check to <postmaster@domain> fails.
4283 SC/01 Eximstats: added -xls and the ability to specify output files
4284 (patch written by Frank Heydlauf).
4286 SC/02 Eximstats: use FileHandles for outputting results.
4288 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4290 SC/04 Eximstats: fixed display of large numbers with -nvr option
4292 SC/05 Eximstats: fixed merging of reports with empty tables.
4294 SC/06 Eximstats: added the -include_original_destination flag
4296 SC/07 Eximstats: removed tabs and trailing whitespace.
4298 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4300 TK/06 MBOX spool code: Add real "From " MBOX separator line
4301 so the .eml file is really in mbox format (even though
4302 most programs do not really care). Patch from Alex Miller.
4304 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4305 The latter is generated from $received_to and is only set if the
4306 message has one envelope recipient. SA can use these headers,
4307 obviously out-of-the-box. Patch from Alex Miller.
4309 PH/08 The ${def test on a variable was returning false if the variable's
4310 value was "0", contrary to what the specification has always said!
4311 The result should be true unless the variable is empty.
4313 PH/09 The syntax error of a character other than { following "${if
4314 def:variable_name" (after optional whitespace) was not being diagnosed.
4315 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4316 accidental colon was present, for example, could give incorrect results.
4318 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4319 result is used (not including appendfile, where other changes are about
4322 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4323 This involved changing a lot of size variables from int to off_t. It
4324 should work with maildirs and everything.
4326 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4327 spamd dying while we are connected to it.
4329 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4332 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4333 function for BDB 4.3.
4335 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4337 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4338 the sources for PCRE have been re-arranged and more files are now
4341 PH/15 The code I had for printing potentially long long variables in PH/11
4342 above was not the best (it lost precision). The length of off_t variables
4343 is now inspected at build time, and an appropriate printing format (%ld
4344 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4345 to be "long long int" or "long int". This is needed for the internal
4346 formatting function string_vformat().
4348 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4349 the configuration file to be ":syslog", then the script "guesses" where
4350 the logs files are, rather than using the compiled in default. In our
4351 case the guess is not the same as the compiled default, so the script
4352 suddenly stopped working when I started to use syslog. The patch checks
4353 to see if log_file_path is "". If so, it attempts to read it from exim
4354 with no configuration file to get the compiled in version, before it
4355 falls back to the previous guessing code."
4357 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4358 implementing BATV in an Exim configuration. See NewStuff for the gory
4361 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4362 Makefile that are specific to HP-UX.
4364 PH/18 If the "use_postmaster" option was set for a recipient callout together
4365 with the "random" option, the postmaster address was used as the MAIL
4366 FROM address for the random test, but not for the subsequent recipient
4367 test. It is now used for both.
4369 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4370 patch removes a few documentation additions to RFC 3028, because the
4371 latest draft now contains them. It adds the new en;ascii-case comparator
4372 and a new error check for 8bit text in MIME parts. Comparator and
4373 require names are now matched exactly. I enabled the subaddress
4374 extension, but it is not well tested yet (read: it works for me)."
4376 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4377 rework some of the code of TK/09 above to avoid the hardwired use of
4378 "%lld" and "long long". Replaced the call to snprintf() with a call to
4381 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4382 records point to non-existent hosts", "retry timeout exceeded", and
4383 "retry time not reached for any host after a long failure period".
4385 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4386 experimental DomainKeys support:
4388 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4389 (2) On an error such as an illegally used "control", the wrong name for
4390 the control was given.
4392 These problems did NOT occur unless DomainKeys support was compiled.
4394 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4396 PH/24 Added ${if match_ip condition.
4398 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4399 databases so that it will be absolutely obvious if a crash occurs in the
4400 DB library. This is a regular occurrence (often caused by mis-matched
4403 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4404 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4405 picked up on a user's system that detects such things. There doesn't seem
4406 to be a gcc warning option for this - only an attribute that has to be
4407 put on the function's prototype. It seems that in Fedora Core 4 they have
4408 set this on a number of new functions. No doubt there will be more in due
4411 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4412 specify a transport. However, if an address that was verified by such a
4413 router was the subject of a callout, Exim crashed because it tried to
4414 read the rcpt_include_affixes from the non-existent transport. Now it
4415 just assumes that the setting of that option is false. This bug was
4416 introduced by 4.51/PH/31.
4418 PH/28 Changed -d+all to exclude +memory, because that information is very
4419 rarely of interest, but it makes the output a lot bigger. People tend to
4420 do -d+all out of habit.
4422 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4423 code in os-type was giving problems when libc.so lives in lib64, like on
4426 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4427 aren't the modern standard, and it seems that some systems' include files
4428 don't always have them. Exim was already checking for some of the newer
4429 ones like T_AAAA, and defining it itself. I've added checks for all the
4430 record types that Exim uses.
4432 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4433 not automatically generating a new one, as it is supposed to. This
4434 prevented TLS from working. If the file did exist, but contained invalid
4435 data, a new version was generated, as expected. It was only the case of a
4436 non-existent file that was broken.
4438 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4439 with a change in libdomainkeys > 0.64.
4441 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4442 from DNS. If the selector record carries the flag, it now has
4443 precedence over the domain-wide flag.
4445 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4447 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4448 the use of a port name that isn't defined in /etc/services) occurred, the
4449 message was deferred as in a normal delivery, and thus remained on the
4450 spool, instead of being failed because of the mua_wrapper setting. This
4451 is now fixed, and I tidied up some of the mua_wrapper messages at the
4454 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4455 the messages in a hash of arrays rather than using individual hashes.
4456 This is a bit cleaner and results in dramatic memory savings, albeit
4457 at a slight CPU cost.
4459 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4460 as requested by Marc Sherman.
4462 SC/10 Eximstats: added histograms for user specified patterns as requested
4465 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4467 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4468 fopen() in the content-scanning modules that did not already have it.
4474 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4475 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4477 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4479 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4481 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4482 to test Sieve filters that use "vacation".
4484 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4485 that changes the way the GnuTLS parameters are stored in the cache file.
4486 The new format can be generated externally. For backward compatibility,
4487 if the data in the cache doesn't make sense, Exim assumes it has read an
4488 old-format file, and it generates new data and writes a new file. This
4489 means that you can't go back to an older release without removing the
4492 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4493 work if there are any delivery delays because "one_time" forces the
4494 parent to be marked "delivered", so its unseen clone is never tried
4495 again. For this reason, Exim now forbids the simultaneous setting of
4498 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4499 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4500 entry if you want to know the details. The fix, however, neglected to
4501 consider the case when local delivery batching is involved. The test for
4502 "previously delivered" was not happening when checking to see if an
4503 address could be batched with a previous (undelivered) one; under
4504 certain circumstances this could lead to multiple deliveries to the same
4507 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4508 in its include files, and this causes problems building Exim.
4510 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4511 header_syntax) but Exim was just ignoring anything given after a slash.
4512 In particular, this caused confusion with an attempt to use "verify =
4513 reverse_host_lookup/defer_ok". An error is now given when options are
4514 supplied for verify items that do not have them. (Maybe reverse_host_
4515 lookup should have a defer_ok option, but that's a different point.)
4517 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4518 defined by RFC 821) to 2048, because there were problems with some AUTH
4519 commands, and RFC 1869 says the size should be increased for extended
4520 SMTP commands that take arguments.
4522 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4525 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4526 "unknown" error; now it says that the functionality isn't in the binary.
4528 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4529 an address' error message when a string expansion fails (syntax or
4530 whatever). Otherwise the password may appear in the log. Following change
4531 PH/42 below, there is no longer a chance of it appearing in a bounce
4534 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4536 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4537 of its addresses was ever tried. (Bugzilla bug #2.)
4539 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4540 the result incorrectly in the debug output. (It correctly added a newline
4541 to what was transported.)
4543 TF/01 Added $received_time.
4545 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4546 commented out examples of how to interface to a virus scanner and to
4547 SpamAssassin. Also added commented examples of av_scanner and
4548 spamd_address settings.
4550 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4551 and controls are allowed in which ACLs. There were a couple of minor
4552 errors. Some of the entries in the conditions table (which is a table of
4553 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4554 negation of where the condition IS allowed.
4556 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4558 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4559 header file does not have a version number, so I've had to invent a new
4560 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4561 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4562 radiusclient), but it was contributed by a Radius user.
4564 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4565 files or directories, for OpenSSL.
4567 PH/19 When an Exim process that is running as root has to create an Exim log
4568 file, it does so in a subprocess that runs as exim:exim so as to get the
4569 ownership right at creation (otherwise, other Exim processes might see
4570 the file with the wrong ownership). There was no test for failure of this
4571 fork() call, which would lead to the process getting stuck as it waited
4572 for a non-existent subprocess. Forks do occasionally fail when resources
4573 run out. I reviewed all the other calls to fork(); they all seem to check
4576 PH/20 When checking for unexpected SMTP input at connect time (before writing
4577 the banner), Exim was not dealing correctly with a non-positive return
4578 from the read() function. If the client had disconnected by this time,
4579 the result was a log entry for a synchronization error with an empty
4580 string after "input=" when read() returned zero. If read() returned -1
4581 (an event I could not check), uninitialized data bytes were printed.
4582 There were reports of junk text (parts of files, etc) appearing after
4585 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4587 PH/22 Added support for macro redefinition, and (re)definition in between
4588 driver and ACL definitions.
4590 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4591 forgetting to use the resulting value; it was using the unexpanded value.
4593 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4594 hadn't been configured. The fix is from Juergen Kreileder, who
4595 understands it better than I do:
4597 "Here's what I see happening with three configured cyrus_sasl
4598 authenticators configured (plain, login, cram-md5):
4600 On startup auth_cyrus_sasl_init() gets called for each of these.
4601 This means three calls to sasl_listmech() without a specified mech_list.
4602 => SASL tests which mechs of all available mechs actually work
4603 => three warnings about OTP not working
4604 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4606 With the patch, sasl_listmech() also gets called three times. But now
4607 SASL's mech_list option is set to the server_mech specified in the the
4608 authenticator. Or in other words, the answer from sasl_listmech()
4609 gets limited to just the mech you're testing for (which is different
4611 => the return list contains just 'plain' or 'login', 'cram-md5' or
4612 nothing depending on the value of ob->server_mech.
4614 I've just tested the patch: Authentication still works fine,
4615 unavailable mechs specified in the exim configuration are still
4616 caught, and the auth.log warnings about OTP are gone."
4618 PH/25 When debugging is enabled, the contents of the command line are added
4619 to the debugging output, even when log_selector=+arguments is not
4622 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4623 answer is "GNU", and only if the return is "GNU/something" is the answer
4626 PH/27 $acl_verify_message is now set immediately after the failure of a
4627 verification in an ACL, and so is available in subsequent modifiers. In
4628 particular, the message can be preserved by coding like this:
4630 warn !verify = sender
4631 set acl_m0 = $acl_verify_message
4633 Previously, $acl_verify_message was set only while expanding "message"
4634 and "log_message" when a very denied access.
4636 PH/28 Modified OS/os.c-Linux with
4638 -#ifndef OS_LOAD_AVERAGE
4639 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4641 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4642 nomenclature these days.)
4644 PH/29 Installed patch from the Sieve maintainer that adds the options
4645 sieve_useraddress and sieve_subaddress to the redirect router.
4647 PH/30 In these circumstances:
4648 . Two addresses routed to the same list of hosts;
4649 . First host does not offer TLS;
4650 . First host accepts first address;
4651 . First host gives temporary error to second address;
4652 . Second host offers TLS and a TLS session is established;
4653 . Second host accepts second address.
4654 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4655 and peerdn, if requested) that were in fact used only for the second
4658 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4659 attention to any local part prefix or suffix that was matched by the
4660 router that accepted the address. It now behaves in the same way as it
4661 does for delivery: the affixes are removed from the local part unless
4662 rcpt_include_affixes is set on the transport.
4664 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4665 timeout during the DATA phase of an incoming message.
4667 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4668 applied a patch sanctioned by the Sieve maintainer.
4670 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4671 the uid or gid is negative. A case of a negative gid caused this to be
4672 noticed. The fix allows for either to be negative.
4674 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4675 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4678 PH/36 The change PH/12 above was broken. Fixed it.
4680 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4681 the grounds that routing the same address twice would always produce the
4682 same answer. This might have been true once, but it is certainly no
4683 longer true now. Routing a child address may depend on the previous
4684 routing that produced that child. Some complicated redirection strategies
4685 went wrong when messages had multiple recipients, and made Exim's
4686 behaviour dependent on the order in which the addresses were given.
4688 I have moved the duplicate checking until after the routing is complete.
4689 Exim scans the addresses that are assigned to local and remote
4690 transports, and removes any duplicates. This means that more work will be
4691 done, as duplicates will always all be routed, but duplicates are
4692 presumably rare, so I don't expect this is of any significance.
4694 For deliveries to pipes, files, and autoreplies, the duplicate checking
4695 still happens during the routing process, since they are not going to be
4698 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4699 It corrects a timeout issue with spamd. This is Ian's comment: "The
4700 background is that sometimes spamd either never reads data from a
4701 connection it has accepted, or it never writes response data. The exiscan
4702 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4703 blindly assumes that writes won't block so it may never time out."
4705 PH/39 Allow G after quota size as well as K and M.
4707 PH/40 The value set for $authenticated_id in an authenticator may not contain
4708 binary zeroes or newlines because the value is written to log lines and
4709 to spool files. There was no check on this. Now the value is run through
4710 the string_printing() function so that such characters are converted to
4711 printable escape sequences.
4713 PH/41 $message_linecount is a new variable that contains the total number of
4714 lines in the message. Compare $body_linecount, which is the count for the
4717 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4718 bounce and delay warning messages, except in certain special cases, which
4721 (a) An SMTP error message from a remote host;
4722 (b) A message specified in a :fail: redirection;
4723 (c) A message specified in a "fail" command in a system filter;
4724 (d) A message specified in a FAIL return from the queryprogram router;
4725 (e) A message specified by the cannot_route_message router option.
4727 In these cases only, Exim does include the error details in bounce and
4728 warning messages. There are also a few cases where bland messages such
4729 as "unrouteable address" or "local delivery error" are given.
4731 PH/43 $value is now also set for the "else" part of a ${run expansion.
4733 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4734 being worked on, but at least Exim now implements the latest version to
4737 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4738 process to complete was treated as a delivery failure, a timeout while
4739 writing the message to the pipe was logged, but erroneously treated as a
4740 successful delivery. Such timeouts include transport filter timeouts. For
4741 consistency with the overall process timeout, these timeouts are now
4742 treated as errors, giving rise to delivery failures by default. However,
4743 there is now a new Boolean option for the pipe transport called
4744 timeout_defer, which, if set TRUE, converts the failures into defers for
4745 both kinds of timeout. A transport filter timeout is now identified in
4748 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4749 systems where "make" and "gmake" are different, calling "gmake" at top
4750 level broke things. I've arranged for the value of $(MAKE) to be passed
4751 from the Makefile to this script so that it can call the same version of
4755 A note about Exim versions 4.44 and 4.50
4756 ----------------------------------------
4758 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4759 changes of various kinds. As a consequence, a big documentation update was
4760 needed. This delayed the release for rather longer than seemed good, especially
4761 in the light of a couple of (minor) security issues. Therefore, the changes
4762 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4763 release. So 4.44 and 4.50 are in effect two different branches that both start
4766 I have left the 4.50 change log unchanged; it contains all the changes since
4767 4.43. The change log for 4.44 is below; many of its items are identical to
4768 those for 4.50. This seems to be the most sensible way to preserve the
4769 historical information.
4775 1. Minor wording change to the doc/README.SIEVE file.
4777 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4778 computation of the current number of files was incorrect.
4780 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4781 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4784 4. Give more explanation in the error message when the command for a transport
4785 filter fails to execute.
4787 5. There are several places where Exim runs a non-Exim command in a
4788 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4789 was being done only for the command run by the queryprogram router. It is
4790 now done for all such subprocesses. The other cases are: ${run, transport
4791 filters, and the commands run by the lmtp and pipe transports.
4793 6. Added CONFIGURE_GROUP build-time option.
4795 7. Some older OS have a limit of 256 on the maximum number of file
4796 descriptors. Exim was using setrlimit() to set 1000 as a large value
4797 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4798 systems. I've change it so that if it can't get 1000, it tries for 256.
4800 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4801 was an oversight, and furthermore, ever since the addition of extra
4802 controls (e.g. 4.43/32), the checks on when to allow different forms of
4803 "control" were broken. There should now be diagnostics for all cases when a
4804 control that does not make sense is encountered.
4806 9. Added the /retain_sender option to "control=submission".
4808 10. $recipients is now available in the predata ACL (oversight).
4810 11. Tidy the search cache before the fork to do a delivery from a message
4811 received from the command line. Otherwise the child will trigger a lookup
4812 failure and thereby defer the delivery if it tries to use (for example) a
4813 cached ldap connection that the parent has called unbind on.
4815 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4816 of $address_data from the recipient verification was clobbered by the
4817 sender verification.
4819 13. The value of address_data from a sender verification is now available in
4820 $sender_address_data in subsequent conditions in the ACL statement.
4822 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4824 15. Added a new option "connect=<time>" to callout options, to set a different
4827 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4828 was its contents. (It was OK if the option was not defined at all.)
4830 17. A "Completed" log line is now written for messages that are removed from
4831 the spool by the -Mrm option.
4833 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4834 information about exactly what failed.
4836 19. Added -dd to debug only the daemon process.
4838 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4839 handles timeouts, both on the server side and network timeouts. Renamed the
4840 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4842 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4843 It is now set to "smtps".
4845 22. $host_address is now set to the target address during the checking of
4846 ignore_target_hosts.
4848 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4849 being passed; this would have caused $sender_host_name to have been used if
4850 matching the list had actually called for a host name (not very likely,
4851 since this list is usually IP addresses). A host name is now passed as
4854 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4855 code that specifies a non-restarting handler (typically sigaction() in
4856 modern systems) in an attempt to fix a rare and obscure crash bug.
4858 25. Narrowed the window for a race in the daemon that could cause it to ignore
4859 SIGCHLD signals. This is not a major problem, because they are used only to
4860 wake it up if nothing else does.
4862 26. A malformed maildirsize file could cause Exim to calculate negative values
4863 for the mailbox size or file count. Odd effects could occur as a result.
4864 The maildirsize information is now recalculated if the size or filecount
4867 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4868 support for a long time. Removed HAVE_SYS_VFS_H.
4870 28. Installed the latest version of exipick from John Jetmore.
4872 29. In an address list, if the pattern was not a regular expression, an empty
4873 subject address (from a bounce message) matched only if the pattern was an
4874 empty string. Non-empty patterns were not even tested. This was the wrong
4875 because it is perfectly reasonable to use an empty address as part of a
4876 database query. An empty address is now tested by patterns that are
4877 lookups. However, all the other forms of pattern expect the subject to
4878 contain a local part and a domain, and therefore, for them, an empty
4879 address still always fails if the pattern is not itself empty.
4881 30. Exim went into a mad DNS loop when attempting to do a callout where the
4882 host was specified on an smtp transport, and looking it up yielded more
4883 than one IP address.
4885 31. Re-factored the code for checking spool and log partition space into a
4886 function that finds that data and another that does the check. The former
4887 is then used to implement four new variables: $spool_space, $log_space,
4888 $spool_inodes, and $log_inodes.
4890 32. The RFC2047 encoding function was originally intended for short strings
4891 such as real names; it was not keeping to the 75-character limit for
4892 encoded words that the RFC imposes. It now respects the limit, and
4893 generates multiple encoded words if necessary. To be on the safe side, I
4894 have increased the buffer size for the ${rfc2047: expansion operator from
4897 33. It is now permitted to omit both strings after an "if" condition; if the
4898 condition is true, the result is "true". As before, when the second string
4899 is omitted, a false condition yields an empty string. This makes it less
4900 cumbersome to write custom ACL and router conditions.
4902 34. Failure to deliver a bounce message always caused it to be frozen, even if
4903 there was an errors_to setting on the router. The errors_to setting is now
4906 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4907 canonical form (fully expanded) before being placed in
4908 $sender_host_address.
4910 36. The table in the code that translates DNS record types into text (T_A to
4911 "A" for instance) was missing entries for NS and CNAME. It is just possible
4912 that this could have caused confusion if both these types were looked up
4913 for the same domain, because the text type is used as part of Exim's
4914 per-process caching. But the chance of anyone hitting this buglet seems
4917 37. The dnsdb lookup has been extended in a number of ways.
4919 (1) There is a new type, "zns", which walks up the domain tree until it
4920 finds some nameserver records. It should be used with care.
4922 (2) There is a new type, "mxh", which is like "mx" except that it returns
4923 just the host names, not the priorities.
4925 (3) It is now possible to give a list of domains (or IP addresses) to be
4926 looked up. The behaviour when one of the lookups defers can be
4927 controlled by a keyword.
4929 (4) It is now possible to specify the separator character for use when
4930 multiple records are returned.
4932 38. The dnslists ACL condition has been extended: it is now possible to supply
4933 a list of IP addresses and/or domains to be looked up in a particular DNS
4936 39. Added log_selector=+queue_time_overall.
4938 40. When running the queue in the test harness, wait just a tad after forking a
4939 delivery process, to get repeatability of debugging output.
4941 41. Include certificate and key file names in error message when GnuTLS fails
4942 to set them up, because the GnuTLS error message doesn't include the name
4943 of the failing file when there is a problem reading it.
4945 42. Allow both -bf and -bF in the same test run.
4947 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4949 44. The "Exiscan patch" is now merged into the mainline Exim source.
4951 45. Sometimes the final signoff response after QUIT could fail to get
4952 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4953 < 0 before doing a fflush(). This bug looks as though it goes back to the
4954 introduction of TLS in release 3.20, but "sometimes" must have been rare
4955 because the tests only now provoked it.
4957 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4958 (this can affect the format of dates).
4960 47. exim_tidydb, when checking for the continued existence of a message for
4961 which it has found a message-specific retry record, was not finding
4962 messages that were in split spool directories. Consequently, it was
4963 deleting retry records that should have stayed in existence.
4965 48. Steve fixed some bugs in eximstats.
4967 49. The SPA authentication driver was not abandoning authentication and moving
4968 on to the next authenticator when an expansion was forced to fail,
4969 contradicting the general specification for all authenticators. Instead it
4970 was generating a temporary error. It now behaves as specified.
4972 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4973 (the order specifies the preference for clients). The order is now AES256,
4974 AES128, 3DES, ARCFOUR128.
4976 51. Small patch to Sieve code - explicitly set From: when generating an
4979 52. Exim crashed if a remote delivery caused a very long error message to be
4980 recorded - for instance if somebody sent an entire SpamAssassin report back
4981 as a large number of 550 error lines. This bug was coincidentally fixed by
4982 increasing the size of one of Exim's internal buffers (big_buffer) that
4983 happened as part of the Exiscan merge. However, to be on the safe side, I
4984 have made the code more robust (and fixed the comments that describe what
4987 53. Now that there can be additional text after "Completed" in log lines (if
4988 the queue_time_overall log selector is set), a one-byte patch to exigrep
4989 was needed to allow it to recognize "Completed" as not the last thing in
4992 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4993 patch that reportedly fixes this has been added. I am not expert enough to
4994 create a test for it. This is what the patch creator wrote:
4996 "I found a little strange behaviour of ldap code when working with
4997 Windows 2003 AD Domain, where users was placed in more than one
4998 Organization Units. When I tried to give exim partial DN, the exit code
4999 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5000 But simultaneously result of request was absolutely normal ldap result,
5001 so I produce this patch..."
5003 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5004 REFERENCE, so I have modified the code to exclude the patch when that macro
5007 55. Some experimental protocols are using DNS PTR records for new purposes. The
5008 keys for these records are domain names, not reversed IP addresses. The
5009 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5010 leaves it alone. Component reversal etc. now happens only for IP addresses.
5013 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5015 57. Double the size of the debug message buffer (to 2048) so that more of very
5016 long debug lines gets shown.
5018 58. The exicyclog utility now does better if the number of log files to keep
5019 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5021 59. Two changes related to the smtp_active_hostname option:
5023 (1) $smtp_active_hostname is now available as a variable.
5024 (2) The default for smtp_banner uses $smtp_active_hostname instead
5025 of $primary_hostname.
5027 60. The host_aton() function is supposed to be passed a string that is known
5028 to be a valid IP address. However, in the case of IPv6 addresses, it was
5029 not checking this. This is a hostage to fortune. Exim now panics and dies
5030 if the condition is not met. A case was found where this could be provoked
5031 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5032 components; fortuitously, this particular loophole had already been fixed
5033 by change 4.50/55 above.
5035 If there are any other similar loopholes, the new check in host_aton()
5036 itself should stop them being exploited. The report I received stated that
5037 data on the command line could provoke the exploit when Exim was running as
5038 exim, but did not say which command line option was involved. All I could
5039 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5040 running as the user.
5043 61. There was a buffer overflow vulnerability in the SPA authentication code
5044 (which came originally from the Samba project). I have added a test to the
5045 spa_base64_to_bits() function which I hope fixes it.
5048 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5049 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5051 63. The daemon start-up calls getloadavg() while still root for those OS that
5052 need the first call to be done as root, but it missed one case: when
5053 deliver_queue_load_max is set with deliver_drop_privilege. This is
5054 necessary for the benefit of the queue runner, because there is no re-exec
5055 when deliver_drop_privilege is set.
5057 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5058 This has been fixed.
5060 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5061 was in use, was not putting the data itself into the right store pool;
5062 consequently, it could be overwritten for a subsequent message in the same
5063 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5066 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5068 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5069 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5070 the function were treating the return as a boolean value, which happened to
5071 work because 0=false and not-0=true, but is not correct code.
5073 68. The host_aton() function was not handling scoped IPv6 addresses (those
5074 with, for example, "%eth0" on the end) correctly.
5076 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5077 negated items (that is, ~something) in unsigned ints. Some compilers
5078 apparently mutter when there is no cast.
5080 70. If an address verification called from an ACL failed, and did not produce a
5081 user-specific message (i.e. there was only a "system" message), nothing was
5082 put in $acl_verify_message. In this situation, it now puts the system
5085 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5086 session; change 4.31/43 added the unwanted input to the log line - except
5087 that it did not do this in the start of session case. It now does.
5089 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5090 This is wrong and can cause the other end to generate a synchronization
5091 error if it is another Exim or anything else that does the synchronization
5092 check. A QUIT command is no longer sent after a timeout.
5094 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5095 during host lookups.
5097 74. The defer_ok option of callout verification was not working if it was used
5098 when verifying addresses in header lines, that is, for this case:
5100 verify = header_sender/callout=defer_ok
5102 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5103 those file descriptors could be used for SMTP connections. If anything
5104 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5105 could be sent to the SMTP client, causing chaos. The daemon now opens
5106 stdin, stdout, and stderr to /dev/null when it puts itself into the
5109 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5110 log by default. The user can override this with suitable Perl magic.
5112 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5113 the log message when discard triggers, was not working for the DATA ACL or
5114 for the non-SMTP ACL.
5116 78. Error message wording change in sieve.c.
5118 79. If smtp_accept_max_per_host was set, the number of connections could be
5119 restricted to fewer than expected, because the daemon was trying to set up
5120 a new connection before checking whether the processes handling previous
5121 connections had finished. The check for completed processes is now done
5122 earlier. On busy systems, this bug wouldn't be noticed because something
5123 else would have woken the daemon, and it would have reaped the completed
5126 80. If a message was submitted locally by a user whose login name contained one
5127 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5128 It caused a spool format error. I have fixed the spool reading code. A
5129 related problem was that the "from" clause in the Received: line became
5130 illegal because of the space(s). It is now covered by ${quote_local_part.
5132 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5135 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5137 83. Patch from Sieve maintainer to fix unterminated string problem in
5138 "vacation" handling.
5140 84. Some minor changes to the Linux configuration files to help with other
5141 OS variants using glibc.
5143 85. One more patch for Sieve to update vacation handling to latest spec.
5146 ----------------------------------------------------
5147 See the note above about the 4.44 and 4.50 releases.
5148 ----------------------------------------------------
5154 1. Change 4.43/35 introduced a bug that caused file counts to be
5155 incorrectly computed when quota_filecount was set in an appendfile
5158 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5159 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5162 3. Give more explanation in the error message when the command for a transport
5163 filter fails to execute.
5165 4. There are several places where Exim runs a non-Exim command in a
5166 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5167 was being done only for the command run by the queryprogram router. It is
5168 now done for all such subprocesses. The other cases are: ${run, transport
5169 filters, and the commands run by the lmtp and pipe transports.
5171 5. Some older OS have a limit of 256 on the maximum number of file
5172 descriptors. Exim was using setrlimit() to set 1000 as a large value
5173 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5174 systems. I've change it so that if it can't get 1000, it tries for 256.
5176 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5177 was an oversight, and furthermore, ever since the addition of extra
5178 controls (e.g. 4.43/32), the checks on when to allow different forms of
5179 "control" were broken. There should now be diagnostics for all cases when a
5180 control that does not make sense is encountered.
5182 7. $recipients is now available in the predata ACL (oversight).
5184 8. Tidy the search cache before the fork to do a delivery from a message
5185 received from the command line. Otherwise the child will trigger a lookup
5186 failure and thereby defer the delivery if it tries to use (for example) a
5187 cached ldap connection that the parent has called unbind on.
5189 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5190 of $address_data from the recipient verification was clobbered by the
5191 sender verification.
5193 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5194 was its contents. (It was OK if the option was not defined at all.)
5196 11. A "Completed" log line is now written for messages that are removed from
5197 the spool by the -Mrm option.
5199 12. $host_address is now set to the target address during the checking of
5200 ignore_target_hosts.
5202 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5203 being passed; this would have caused $sender_host_name to have been used if
5204 matching the list had actually called for a host name (not very likely,
5205 since this list is usually IP addresses). A host name is now passed as
5208 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5209 code that specifies a non-restarting handler (typically sigaction() in
5210 modern systems) in an attempt to fix a rare and obscure crash bug.
5212 15. Narrowed the window for a race in the daemon that could cause it to ignore
5213 SIGCHLD signals. This is not a major problem, because they are used only to
5214 wake it up if nothing else does.
5216 16. A malformed maildirsize file could cause Exim to calculate negative values
5217 for the mailbox size or file count. Odd effects could occur as a result.
5218 The maildirsize information is now recalculated if the size or filecount
5221 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5222 support for a long time. Removed HAVE_SYS_VFS_H.
5224 18. Updated exipick to current release from John Jetmore.
5226 19. Allow an empty sender to be matched against a lookup in an address list.
5227 Previously the only cases considered were a regular expression, or an
5230 20. Exim went into a mad DNS lookup loop when doing a callout where the
5231 host was specified on the transport, if the DNS lookup yielded more than
5234 21. The RFC2047 encoding function was originally intended for short strings
5235 such as real names; it was not keeping to the 75-character limit for
5236 encoded words that the RFC imposes. It now respects the limit, and
5237 generates multiple encoded words if necessary. To be on the safe side, I
5238 have increased the buffer size for the ${rfc2047: expansion operator from
5241 22. Failure to deliver a bounce message always caused it to be frozen, even if
5242 there was an errors_to setting on the router. The errors_to setting is now
5245 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5246 canonical form (fully expanded) before being placed in
5247 $sender_host_address.
5249 24. Updated eximstats to version 1.33
5251 25. Include certificate and key file names in error message when GnuTLS fails
5252 to set them up, because the GnuTLS error message doesn't include the name
5253 of the failing file when there is a problem reading it.
5255 26. Expand error message when OpenSSL has problems setting up cert/key files.
5258 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5259 (this can affect the format of dates).
5261 28. exim_tidydb, when checking for the continued existence of a message for
5262 which it has found a message-specific retry record, was not finding
5263 messages that were in split spool directories. Consequently, it was
5264 deleting retry records that should have stayed in existence.
5266 29. eximstats updated to version 1.35
5267 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5268 1.35 - bugfix such that pie charts by volume are generated correctly
5270 30. The SPA authentication driver was not abandoning authentication and moving
5271 on to the next authenticator when an expansion was forced to fail,
5272 contradicting the general specification for all authenticators. Instead it
5273 was generating a temporary error. It now behaves as specified.
5275 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5276 (the order specifies the preference for clients). The order is now AES256,
5277 AES128, 3DES, ARCFOUR128.
5279 31. Small patch to Sieve code - explicitly set From: when generating an
5282 32. Exim crashed if a remote delivery caused a very long error message to be
5283 recorded - for instance if somebody sent an entire SpamAssassin report back
5284 as a large number of 550 error lines. This bug was coincidentally fixed by
5285 increasing the size of one of Exim's internal buffers (big_buffer) that
5286 happened as part of the Exiscan merge. However, to be on the safe side, I
5287 have made the code more robust (and fixed the comments that describe what
5290 33. Some experimental protocols are using DNS PTR records for new purposes. The
5291 keys for these records are domain names, not reversed IP addresses. The
5292 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5293 leaves it alone. Component reversal etc. now happens only for IP addresses.
5296 34. The host_aton() function is supposed to be passed a string that is known
5297 to be a valid IP address. However, in the case of IPv6 addresses, it was
5298 not checking this. This is a hostage to fortune. Exim now panics and dies
5299 if the condition is not met. A case was found where this could be provoked
5300 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5301 components; fortuitously, this particular loophole had already been fixed
5302 by change 4.50/55 or 4.44/33 above.
5304 If there are any other similar loopholes, the new check in host_aton()
5305 itself should stop them being exploited. The report I received stated that
5306 data on the command line could provoke the exploit when Exim was running as
5307 exim, but did not say which command line option was involved. All I could
5308 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5309 running as the user.
5312 35. There was a buffer overflow vulnerability in the SPA authentication code
5313 (which came originally from the Samba project). I have added a test to the
5314 spa_base64_to_bits() function which I hope fixes it.
5317 36. The daemon start-up calls getloadavg() while still root for those OS that
5318 need the first call to be done as root, but it missed one case: when
5319 deliver_queue_load_max is set with deliver_drop_privilege. This is
5320 necessary for the benefit of the queue runner, because there is no re-exec
5321 when deliver_drop_privilege is set.
5323 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5324 was in use, was not putting the data itself into the right store pool;
5325 consequently, it could be overwritten for a subsequent message in the same
5326 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5329 38. Sometimes the final signoff response after QUIT could fail to get
5330 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5331 < 0 before doing a fflush(). This bug looks as though it goes back to the
5332 introduction of TLS in release 3.20, but "sometimes" must have been rare
5333 because the tests only now provoked it.
5339 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5340 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5341 Now it returns an integer. A number of calls were still expecting a T/F
5342 return. Fortuitously, in all cases, the tests worked in OK situations,
5343 which is the norm. However, things would have gone wrong on any write
5344 failures on the smtp file descriptor. This function is used when sending
5345 messages over SMTP and also when doing verify callouts.
5347 2. When Exim is called to do synchronous delivery of a locally submitted
5348 message (the -odf or -odi options), it no longer closes stderr before doing
5351 3. Implemented the mua_wrapper option.
5353 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5355 5. Implemented the functions header_remove(), header_testname(),
5356 header_add_at_position(), and receive_remove_recipient(), and exported them
5359 6. If an ACL "warn" statement specified the addition of headers, Exim already
5360 inserted X-ACL-Warn: at the start if there was no header name. However, it
5361 was not making this test for the second and subsequent header lines if
5362 there were newlines in the string. This meant that an invalid header could
5363 be inserted if Exim was badly configured.
5365 7. Allow an ACL "warn" statement to add header lines at the start or after all
5366 the Received: headers, as well as at the end.
5368 8. Added the rcpt_4xx retry error code.
5370 9. Added postmaster_mailfrom=xxx to callout verification option.
5372 10. Added mailfrom=xxxx to the callout verification option, for verify=
5375 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5376 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5378 12. Inserted some casts to stop certain compilers warning when using pointer
5379 differences as field lengths or precisions in printf-type calls (mostly
5380 affecting debugging statements).
5382 13. Added optional readline() support for -be (dynamically loaded).
5384 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5385 same clock tick as a message's arrival, so that its received time was the
5386 same as the "first fail" time on the retry record, and that message
5387 remained on the queue past the ultimate address timeout, every queue runner
5388 would try a delivery (because it was past the ultimate address timeout) but
5389 after another failure, the ultimate address timeout, which should have then
5390 bounced the address, did not kick in. This was a "< instead of <=" error;
5391 in most cases the first failure would have been in the next clock tick
5392 after the received time, and all would be well.
5394 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5395 being recognized when the domain list was tested by the match_domain
5396 condition in an expansion string.
5398 16. Added the ${str2b64: operator.
5400 17. Exim was always calling setrlimit() to set a large limit for the number of
5401 processes, without checking whether the existing limit was already
5402 adequate. (It did check for the limit on file descriptors.) Furthermore,
5403 errors from getrlimit() and setrlimit() were being ignored. Now they are
5404 logged to the main and panic logs, but Exim does carry on, to try to do its
5405 job under whatever limits there are.
5407 18. Imported PCRE 5.0.
5409 19. Trivial typo in log message " temporarily refused connection" (the leading
5412 20. If the log selector return_path_on_delivery was set and an address was
5413 redirected to /dev/null, the delivery process crashed because it assumed
5414 that a return path would always be set for a "successful" delivery. In this
5415 case, the whole delivery is bypassed as an optimization, and therefore no
5418 21. Internal re-arrangement: the function for sending a challenge and reading
5419 a response while authentication was assuming a zero-terminated challenge
5420 string. It's now changed to take a pointer and a length, to allow for
5421 binary data in such strings.
5423 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5425 23. Exim was not respecting finduser_retries when seeking the login of the
5426 uid under which it was called; it was always trying 10 times. (The default
5427 setting of finduser_retries is zero.) Also, it was sleeping after the final
5428 failure, which is pointless.
5430 24. Implemented tls_on_connect_ports.
5432 25. Implemented acl_smtp_predata.
5434 26. If the domain in control=submission is set empty, Exim assumes that the
5435 authenticated id is a complete email address when it generates From: or
5436 Sender: header lines.
5438 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5439 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5440 chown and chgrp in /bin and hostname in /usr/bin.
5442 28. Exim was keeping the "process log" file open after each use, just as it
5443 does for the main log. This opens the possibility of it remaining open for
5444 long periods when the USR1 signal hits a daemon. Occasional processlog
5445 errors were reported, that could have been caused by this. Anyway, it seems
5446 much more sensible not to leave this file open at all, so that is what now
5449 29. The long-running daemon process does not normally write to the log once it
5450 has entered its main loop, and it closes the log before doing so. This is
5451 so that log files can straightforwardly be renamed and moved. However,
5452 there are a couple of unusual error situations where the daemon does write
5453 log entries, and I had neglected to close the log afterwards.
5455 30. The text of an SMTP error response that was received during a remote
5456 delivery was being truncated at 512 bytes. This is too short for some of
5457 the long messages that one sometimes sees. I've increased the limit to
5460 31. It is now possible to make retry rules that apply only when a message has a
5461 specific sender, in particular, an empty sender.
5463 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5464 it possible to be selective about when SMTP synchronization is enforced.
5466 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5468 32. Implemented hosts_connection_nolog.
5470 33. Added an ACL for QUIT.
5472 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5475 35. Added mailbox_size and mailbox_filecount to appendfile.
5477 36. Added control = no_multiline_responses to ACLs.
5479 37. There was a bug in the logic of the code that waits for the clock to tick
5480 in the case where the clock went backwards by a substantial amount such
5481 that the microsecond fraction of "now" was more than the microsecond
5482 fraction of "then" (but the whole seconds number was less).
5484 38. Added support for the libradius Radius client library this is found on
5485 FreeBSD (previously only the radiusclient library was supported).
5491 1. When certain lookups returned multiple values in the form name=value, the
5492 quoting of the values was not always being done properly. Specifically:
5493 (a) If the value started with a double quote, but contained no whitespace,
5495 (b) If the value contained whitespace other than a space character (i.e.
5496 tabs or newlines or carriage returns) it was not quoted.
5497 This fix has been applied to the mysql and pgsql lookups by writing a
5498 separate quoting function and calling it from the lookup code. The fix
5499 should probably also be applied to nisplus, ibase and oracle lookups, but
5500 since I cannot test any of those, I have not disturbed their existing code.
5502 2. A hit in the callout cache for a specific address caused a log line with no
5503 reason for rejecting RCPT. Now it says "Previous (cached) callout
5504 verification failure".
5506 3. There was an off-by-one bug in the queryprogram router. An over-long
5507 return line was truncated at 256 instead of 255 characters, thereby
5508 overflowing its buffer with the terminating zero. As well as fixing this, I
5509 have increased the buffer size to 1024 (and made a note to document this).
5511 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5512 when Exim is waiting for an SMTP response from a remote server, Exim
5513 restarts its select() call on the socket, thereby resetting its timeout.
5514 This is not a problem when such interrupts are rare. Somebody set up a cron
5515 job to run exiwhat every 2 minutes, which is less than the normal select()
5516 timeout (5 or 10 minutes). This meant that the select() timeout never
5517 kicked in because it was always reset. I have fixed this by comparing the
5518 time when an interrupt arrives with the time at the start of the first call
5519 to select(). If more time than the timeout has elapsed, the interrupt is
5520 treated as a timeout.
5522 5. Some internal re-factoring in preparation for the addition of Sieve
5523 extensions (by MH). In particular, the "personal" test is moved to a
5524 separate function, and given an option for scanning Cc: and Bcc: (which is
5525 not set for Exim filters).
5527 6. When Exim created an email address using the login of the caller as the
5528 local part (e.g. when creating a From: or Sender: header line), it was not
5529 quoting the local part when it contained special characters such as @.
5531 7. Installed new OpenBSD configuration files.
5533 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5534 try to make them clearer.
5536 9. Callout options, other than the timeout value, were being ignored when
5537 verifying sender addresses in header lines. For example, when using
5539 verify = header_sender/callout=no_cache
5541 the cache was (incorrectly) being used.
5543 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5544 only the Cygwin environment.
5546 11. When return_path_on_delivery was set as a log selector, if different remote
5547 addresses in the same message used different return paths and parallel
5548 remote delivery occurred, the wrong values would sometimes be logged.
5549 (Whenever a remote delivery process finished, the return path value from
5550 the most recently started remote delivery process was logged.)
5552 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5553 lines when AUTH and/or TLS are in use. This is the "received protocol"
5554 field. Exim used to use "asmtp" for authenticated SMTP, without any
5555 indication (in the protocol name) for TLS use. Now it follows the RFC and
5556 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5557 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5558 names appear in log lines as well as in Received: header lines.
5560 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5561 extensions, and comparison tests, and to fix some bugs.
5563 14. Changes to the "personal" filter test:
5565 (1) The test was buggy in that it was just doing the equivalent of
5566 "contains" tests on header lines. For example, if a user's address was
5567 anne@some.where, the "personal" test would incorrectly be true for
5569 To: susanne@some.where
5571 This test is now done by extracting each address from the header in turn,
5572 and checking the entire address. Other tests that are part of "personal"
5573 are now done using regular expressions (for example, to check local parts
5574 of addresses in From: header lines).
5576 (2) The list of non-personal local parts in From: addresses has been
5577 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5578 taken from the Sieve specification recommendations.
5580 (3) If the message contains any header line starting with "List-" it is
5581 treated as non-personal.
5583 (4) The test for "circular" in the Subject: header line has been removed
5584 because it now seems ill-conceived.
5586 15. Minor typos in src/EDITME comments corrected.
5588 16. Installed latest exipick from John Jetmore.
5590 17. If headers_add on a router specified a text string that was too long for
5591 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5592 of string_sprintf() is now avoided.
5594 18. $message_body_size was not set (it was always zero) when running the DATA
5595 ACL and the local_scan() function.
5597 19. For the "mail" command in an Exim filter, no default was being set for
5598 the once_repeat time, causing a random time value to be used if "once" was
5599 specified. (If the value happened to be <= 0, no repeat happened.) The
5600 default is now 0s, meaning "never repeat". The "vacation" command was OK
5601 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5602 (I found it when inspecting the code).
5604 20. There is now an overall timeout for performing a callout verification. It
5605 defaults to 4 times the callout timeout, which applies to individual SMTP
5606 commands during the callout. The overall timeout applies when there is more
5607 than one host that can be tried. The timeout is checked before trying the
5608 next host. This prevents very long delays if there are a large number of
5609 hosts and all are timing out (e.g. when the network connections are timing
5610 out). The value of the overall timeout can be changed by specifying an
5611 additional sub-option for "callout", called "maxwait". For example:
5613 verify = sender/callout=5s,maxwait=20s
5615 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5616 to the end before writing, but this should make it even safer).
5618 22. Exim was forgetting that it had advertised PIPELINING for the second and
5619 subsequent messages on an SMTP connection. It was also not resetting its
5620 memory on STARTTLS and an internal HELO.
5622 23. When Exim logs an SMTP synchronization error within a session, it now
5623 records whether PIPELINING has been advertised or not.
5625 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5626 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5627 rather than long int.
5629 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5631 26. Added the never_mail option to autoreply.
5637 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5638 crash if the getsockname() call failed; this can happen if a connection is
5639 closed very soon after it is established. The problem was simply in the
5640 order in which certain operations were done, causing Exim to try to write
5641 to the SMTP stream before it had set up the file descriptor. The bug has
5642 been fixed by making things happen in the correct order.
5648 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5649 before the connection was closed, thus losing the rejection response.
5651 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5652 some early Solaris releases, but causes trouble in current releases where
5653 socklen_t is defined.
5655 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5658 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5661 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5662 that releasing the top of it at the end releases what was used for sub-
5663 expansions (unless the block got too big). However, discard this block if
5664 the first thing is a variable or header, so that we can use its block when
5665 it is dynamic (useful for very large $message_headers, for example).
5667 6. Lookups now cache *every* query, not just the most recent. A new, separate
5668 store pool is used for this. It can be recovered when all lookup caches are
5669 flushed. Lookups now release memory at the end of their result strings.
5670 This has involved some general refactoring of the lookup sources.
5672 7. Some code has been added to the store_xxx() functions to reduce the amount
5673 of flapping under certain conditions.
5675 8. log_incoming_interface used to affect only the <= reception log lines. Now
5676 it causes the local interface and port to be added to several more SMTP log
5677 lines, for example "SMTP connection from", and rejection lines.
5679 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5681 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5683 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5684 could be overwritten at the end of the current message (or the start of a
5685 new message if it was set in a HELO ACL). The value is now preserved for
5686 the duration of the SMTP connection.
5688 12. If a transport had a headers_rewrite setting, and a matching header line
5689 contained an unqualified address, that address was qualified, even if it
5690 did not match any rewriting rules. The underlying bug was that the values
5691 of the flags that permit the existence of unqualified sender and recipient
5692 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5693 non-local messages, and by -bnq for local messages) were not being
5694 preserved with the message after it was received.
5696 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5697 "next input=" as part of the text comprising the host identity instead of
5698 the correct text. The code was using the same buffer for two different
5699 strings. However, depending on which order the printing function evaluated
5700 its arguments, the bug did not always show up. Under Linux, for example, my
5701 test suite worked just fine.
5703 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5704 doesn't work with some older versions of Perl. It has been changed to "my",
5705 which in any case is probably the better facility to use.
5707 15. A really picky compiler found some instances of statements for creating
5708 error messages that either had too many or two few arguments for the format
5711 16. The size of the buffer for calls to the DNS resolver has been increased
5712 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5713 for addresses that have a lot of PTR records. This alleviates a problem; it
5714 does not fully solve it.
5716 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5717 buffer now truncates the list and logs the incident, which is the same
5718 action as happens when Exim is looking up a host name and its aliases.
5719 Previously in this situation something unpredictable would happen;
5720 sometimes it was "internal error: store_reset failed".
5722 18. If a server dropped the connection unexpectedly when an Exim client was
5723 using GnuTLS and trying to read a response, the client delivery process
5724 crashed while trying to generate an error log message.
5726 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5727 string, for example:
5729 warn message = H1: something\nH2: something
5731 the text was added as a single header line from Exim's point of view
5732 though it ended up OK in the delivered message. However, searching for the
5733 second and subsequent header lines using $h_h2: did not work. This has been
5734 fixed. Similarly, if a system filter added multiple headers in this way,
5735 the routers could not see them.
5737 20. Expanded the error message when iplsearch is called with an invalid key to
5738 suggest using net-iplsearch in a host list.
5740 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5741 ACLs are no longer actually imposed (and a message to that effect is
5744 22. If a "gecos" field in a passwd entry contained escaped characters, in
5745 particular, if it contained a \" sequence, Exim got it wrong when building
5746 a From: or a Sender: header from that name. A second bug also caused
5747 incorrect handling when an unquoted " was present following a character
5748 that needed quoting.
5750 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5751 was not being matched caselessly.
5753 24. Arranged for all hyphens in the exim.8 source to be escaped with
5756 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5757 properly. Recipient callout cache records were still being keyed to include
5758 the sender, even when use_sender was set false. This led to far more
5759 callouts that were necessary. The sender is no longer included in the key
5760 when use_sender is false.
5762 26. Added "control = submission" modifier to ACLs.
5764 27. Added the ${base62d: operator to decode base 62 numbers.
5766 28. dnsdb lookups can now access SRV records.
5768 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5769 the configuration file.
5771 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5772 (-v) mode. This makes the output for a verbose queue run more intelligible.
5774 31. Added a use_postmaster feature to recipient callouts.
5776 32. Added the $body_zerocount variable, containing the number of binary zero
5777 bytes in the message body.
5779 33. The time of last modification of the "new" subdirectory is now used as the
5780 "mailbox time last read" when there is a quota error for a maildir
5783 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5785 35. Added +ignore_unknown as a special item in host lists.
5787 36. Code for decoding IPv6 addresses in host lists is now included, even if
5788 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5789 address was recognized as an IP address, but was then not correctly decoded
5790 into binary, causing unexpected and incorrect effects when compared with
5797 1. Very minor rewording of debugging text in manualroute to say "list of
5798 hosts" instead of "hostlist".
5800 2. If verify=header_syntax was set, and a header line with an unqualified
5801 address (no domain) and a large number of spaces between the end of the
5802 name and the colon was received, the reception process suffered a buffer
5803 overflow, and (when I tested it) crashed. This was caused by some obsolete
5804 code that should have been removed. The fix is to remove it!
5806 3. When running in the test harness, delay a bit after writing a bounce
5807 message to get a bit more predictability in the log output.
5809 4. Added a call to search_tidyup() just before forking a reception process. In
5810 theory, someone could use a lookup in the expansion of smtp_accept_max_
5811 per_host which, without the tidyup, could leave open a database connection.
5813 5. Added the variables $recipient_data and $sender_data which get set from a
5814 lookup success in an ACL "recipients" or "senders" condition, or a router
5815 "senders" option, similar to $domain_data and $local_part_data.
5817 6. Moved the writing of debug_print from before to after the "senders" test
5820 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5821 problems for message scanning, either using a data ACL, or using
5822 local_scan() because the Received: header was not generated till after they
5823 were called (in order to set the time as the time of reception completion).
5824 I have revised the way this works. The header is now generated after the
5825 body is received, but before the ACL or local_scan() are called. After they
5826 are run, the timestamp in the header is updated.
5832 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5833 before starting a queue runner without re-exec. This happened only when
5834 deliver_drop_privilege was set or when the Exim user was set to root. The
5835 effect of the bug was that timeouts during subsequent deliveries caused
5836 crashes instead of being properly handled. The handler is now left at its
5837 default (and expected) setting.
5839 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5840 message, again when deliver_drop_privilege is set or Exim is run as root.
5841 The bug described in (1) was not present in this case, but the tidying up
5842 of the other signals was missing. I have made the two cases consistent.
5844 3. The ignore_target_hosts setting on a manualroute router was being ignored
5845 for hosts that were looked up using the /MX notation.
5847 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5850 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5851 operated on the sender address. After changing the $sender_address to <>
5852 for the sender address verify, Exim was re-instated it as the original
5853 (before rewriting) address, but remembering that it had rewritten it, so it
5854 wasn't rewriting it again. This bug also had the effect of breaking the
5855 sender address verification caching when the sender address was rewritten.
5857 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5858 This has been changed so that if the ip literal address matches
5859 ignore_target_hosts, the router declines.
5861 7. Added expansion conditions match_domain, match_address, and match_local_
5862 part (NOT match_host).
5864 8. The placeholder for the Received: header didn't have a length field set.
5866 9. Added code to Exim itself and to exim_lock to test for a specific race
5867 condition that could lead to file corruption when using MBX delivery. The
5868 issue is with the lockfile that is created in /tmp. If this file is removed
5869 after a process has opened it but before that process has acquired a lock,
5870 there is the potential for a second process to recreate the file and also
5871 acquire a lock. This could lead to two Exim processes writing to the file
5872 at the same time. The added code performs the same test as UW imapd; it
5873 checks after acquiring the lock that its file descriptor still refers to
5874 the same named file.
5876 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5877 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5880 11. Added the smtp_active_hostname option. If used, this will typically be made
5881 to depend on the incoming interface address. Because $interface_address is
5882 not set up until the daemon has forked a reception process, error responses
5883 that can happen earlier (such as "too many connections") no longer contain
5886 12. If an expansion in a condition on a "warn" statement fails because a lookup
5887 defers, the "warn" statement is abandoned, and the next ACL statement is
5888 processed. Previously this caused the whole ACL to be aborted.
5890 13. Added the iplsearch lookup type.
5892 14. Added ident_timeout as a log selector.
5894 15. Added tls_certificate_verified as a log selector.
5896 16. Added a global option tls_require_ciphers (compare the smtp transport
5897 option of the same name). This controls incoming TLS connections.
5899 17. I finally figured out how to make tls_require_ciphers do a similar thing
5900 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5901 before starting the TLS session.
5903 18. Tabs are now shown as \t in -bP output.
5905 19. If the log selector return_path_on_delivery was set, Exim crashed when
5906 bouncing a message because it had too many Received: header lines.
5908 20. If two routers both had headers_remove settings, and the first one included
5909 a superfluous trailing colon, the final name in the first list and the
5910 first name in the second list were incorrectly joined into one item (with a
5911 colon in the middle).
5917 1. Added -C and -D options to the exinext utility, mainly to make it easier
5918 to include in the automated testing, but these could be helpful when
5919 multiple configurations are in use.
5921 2. The exinext utility was not formatting the output nicely when there was
5922 an alternate port involved in the retry record key, nor when there was a
5923 message id as well (for retries that were specific to a specific message
5924 and a specific host). It was also confused by IPv6 addresses, because of
5925 the additional colons they contain. I have fixed the IPv4 problem, and
5926 patched it up to do a reasonable job for IPv6.
5928 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5929 delivery, the log line now contains "pipelined" if PIPELINING was used.
5931 4. An SMTP transport process used to panic and die if the bind() call to set
5932 an explicit outgoing interface failed. This has been changed; it is now
5933 treated in the same way as a connect() failure.
5935 5. A reference to $sender_host_name in the part of a conditional expansion
5936 that was being skipped was still causing a DNS lookup. This no longer
5939 6. The def: expansion condition was not recognizing references to header lines
5940 that used bh_ and bheader_.
5942 7. Added the _cache feature to named lists.
5944 8. The code for checking quota_filecount in the appendfile transport was
5945 allowing one more file than it should have been.
5947 9. For compatibility with Sendmail, the command line option
5955 and sets the incoming protocol and host name (for trusted callers). The
5956 host name and its colon can be omitted when only the protocol is to be set.
5957 Note the Exim already has two private options, -pd and -ps, that refer to
5958 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5959 "s", but I don't think that's a major issue.
5961 10. A number of refactoring changes to the code, none of which should affect
5964 (a) The number of logging options was getting close to filling up the
5965 32-bit word that was used as a bit map. I have split them into two classes:
5966 those that are passed in the argument to log_write(), and those that are
5967 only ever tested independently outside of that function. These are now in
5968 separate 32-bit words, so there is plenty of room for expansion again.
5969 There is no change in the user interface or the logging behaviour.
5971 (b) When building, for example, log lines, the code previously used a
5972 macro that called string_cat() twice, in order to add two strings. This is
5973 not really sufficiently general. Furthermore, there was one instance where
5974 it was actually wrong because one of the argument was used twice, and in
5975 one call a function was used. (As it happened, calling the function twice
5976 did not affect the overall behaviour.) The macro has been replaced by a
5977 function that can join an arbitrary number of extra strings onto a growing
5980 (c) The code for expansion conditions now uses a table and a binary chop
5981 instead of a serial search (which was left over from when there were very
5982 few conditions). Also, it now recognizes conditions like "pam" even when
5983 the relevant support is not compiled in: a suitably worded error message is
5984 given if an attempt is made to use such a condition.
5986 11. Added ${time_interval:xxxxx}.
5988 12. A bug was causing one of the ddress fields not to be passed back correctly
5989 from remote delivery subprocesses. The field in question was not being
5990 subsequently used, so this caused to problems in practice.
5992 13. Added new log selectors queue_time and deliver_time.
5994 14. Might have fixed a bug in maildirsizefile handling that threw up
5995 "unexpected character" debug warnings, and recalculated the data
5996 unnecessarily. In any case, I expanded the warning message to give more
5999 15. Added the message "Restricted characters in address" to the statements in
6000 the default ACL that block characters like @ and % in local parts.
6002 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6003 Three changes have been made:
6005 (a) There was a serious bug; a negative response to MAIL caused the whole
6006 recipient domain to be cached as invalid, thereby blocking all messages
6007 to all local parts at the same domain, from all senders. This bug has
6008 been fixed. The domain is no longer cached after a negative response to
6009 MAIL if the sender used is not empty.
6011 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6014 (c) A new callout option, "use_sender" has been added for people who want
6015 the modified behaviour.
6021 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6024 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6025 indeed breaks things for older releases.
6027 3. Added additional logging to the case where there is a problem reading data
6028 from a filter that is running in a subprocess using a pipe, in order to
6029 try to track down a specific problem.
6031 4. Testing facility fudge: when running in the test harness and attempting
6032 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6033 getting "No route to host". Convert this to a timeout.
6035 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6038 6. Some OS don't have socklen_t but use size_t instead. This affects the
6039 fifth argument of getsockopt() amongst other things. This is now
6040 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6041 can be set for individual OS. I have set it for SunOS5, OSF1, and
6042 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6043 some earlier ones do not.
6045 7. Change 4.30/15 was not doing the test caselessly.
6047 8. The standard form for an IPv6 address literal was being rejected by address
6048 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6049 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6050 this, as well as the form without the "IPv6" on the front (but only when
6051 address literals are enabled, of course).
6053 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6055 10. Exim crashed if a message with an empty sender address specified by -f
6056 encountered a router with an errors_to setting. This could be provoked only
6057 by a command such as
6061 where an empty string was supplied; "<>" did not hit this bug.
6063 11. Installed PCRE release 4.5.
6065 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6066 remained set. It is now erased.
6068 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6069 times from message ids (which are base 36 rather than the normal 62).
6071 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6072 were being counted as actual protocol errors, and logged if the log
6073 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6074 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6075 and DATA following a set of rejected RCPTs do not count as protocol errors.
6076 In other words, Exim assumes they were pipelined, though this may not
6077 actually be the case. Of course, in all cases the client gets an
6078 appropriate error code.
6080 15. If a lookup fails in an ACL condition, a message about the failure may
6081 be available; it is used if testing the ACL cannot continue, because most
6082 such messages specify what the cause of the deferral is. However, some
6083 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6084 that caused an old message to be retained and used if a later statement
6085 caused a defer, replacing the real cause of the deferral.
6087 16. If an IP address had so many PTR records that the DNS lookup buffer
6088 was not large enough to hold them, Exim could crash while trying to process
6089 the truncated data. It now detects and logs this case.
6091 17. Further to 4.21/58, another change has been made: if (and only if) the
6092 first line of a message (the first header line) ends with CRLF, a bare LF
6093 in a subsequent header line has a space inserted after it, so as not to
6094 terminate the header.
6096 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6097 unnecessarily, used atoi() instead of strtol(), and didn't check the
6098 termination when getting file sizes from file names by regex.
6100 19. Completely re-implemented the support for maildirsize files, in the light
6101 of a number of problems with the previous contributed implementation
6102 (4.30/29). In particular:
6104 . If the quota is zero, the maildirsize file is maintained, but no quota is
6107 . If the maildir directory does not exist, it is created before any attempt
6108 to write a maildirsize file.
6110 . The quota value in the file is just a cache; if the quota is changed in
6111 the transport, the new value overrides.
6113 . A regular expression is available for excluding directories from the
6116 20. The autoreply transport checks the characters in options that define the
6117 message's headers; it allows continued headers, but it was checking with
6118 isspace() after an embedded newline instead of explicitly looking for a
6121 21. If all the "regular" hosts to which an address was routed had passed their
6122 expiry times, and had not reached their retry times, the address was
6123 bounced, even if fallback hosts were defined. Now Exim should go on to try
6126 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6127 equivalent code in the SMTP transport. Some hosts send humungous responses
6128 to HELO/EHLO, more than 1024 it seems.
6130 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6131 gives compiler warnings in some environments. I've now done it "properly",
6134 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6135 (because of problems with the built-in one) was declared to return uschar *
6136 instead of char *, causing compiler failure.
6138 25. Fixed a file descriptor leak when processing alias/forward files.
6140 26. Fixed a minor format string issue in dbfn.c.
6142 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6144 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6145 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6148 29. When a sender address is verified, it is cached, to save repeating the test
6149 when there is more than one recipient in a message. However, when the
6150 verification involves a callout, it is possible for different callout
6151 options to be set for different recipients. It is too complicated to keep
6152 track of this in the cache, so now Exim always runs a verification when a
6153 callout is required, relying on the callout cache for the optimization.
6154 The overhead is duplication of the address routing, but this should not be
6157 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6158 to be verified with callout=postmaster, and the main callout worked but the
6159 postmaster check failed, the verification correctly failed. However, if a
6160 subsequent RCPT command asked for sender verification *without* the
6161 postmaster check, incorrect caching caused this verification also to fail,
6164 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6165 it was not caching the DNS options (qualify_single, search_parents) that
6166 were used when the lookup failed. A subsequent lookup with different
6167 options therefore always gave the same answer, though there were cases
6168 where it should not have. (Example: a "domains = !$mx_any" option on a
6169 dnslookup router: the "domains" option is always processed without any
6170 widening, but the router might have qualify_single set.) Now Exim uses the
6171 cached value only when the same options are set.
6173 32. Added John Jetmore's "exipick" utility to the distribution.
6175 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6176 than a timeout (e.g. a certificate is required, and is not provided), an
6177 Exim server now closes the connection immediately. Previously it waited for
6178 the client to close - but if the client is SSL, it seems that they each
6179 wait for each other, leading to a delay before one of them times out.
6181 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6182 maintained 0.8.x compatibility because I don't think many are using it, and
6183 it is clearly obsolete.
6185 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6188 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6189 client certificate was expired. A simple patch fixes this, though I don't
6190 understand the full logic of why the verify callback is called multiple
6193 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6194 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6195 which causes problems with some clients (such as the Certicom SSL Plus
6196 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6197 disables the coutermeasure allowing Eudora to connect."
6199 38. Exim was not checking that a write() to a log file succeeded. This could
6200 lead to Bad Things if a log got too big, in particular if it hit a file
6201 size limit. Exim now panics and dies if it cannot write to a log file, just
6202 as it does if it cannot open a log file.
6204 39. Modified OS/Makefile-Linux so that it now contains
6206 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6208 The two -D definitions ensure that Exim is compiled with large file
6209 support, which makes it possible to handle log files that are bigger than
6212 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6213 instance) a domain was checked against a named list that involved a lookup,
6214 causing $domain_data to be set, then another domain was checked against the
6215 same list, then the first domain was re-checked, the value of $domain_data
6216 after the final check could be wrong. In particular, if the second check
6217 failed, it could be set empty. This bug probably also applied to
6220 41. The strip_trailing_dot option was not being applied to the address given
6221 with the -f command-line option.
6223 42. The code for reading a message's header from the spool was incrementing
6224 $received_count, but never initializing it. This meant that the value was
6225 incorrect (doubled) while delivering a message in the same process in which
6226 it was received. In the most common configuration of Exim, this never
6227 happens - a fresh exec is done - but it can happen when
6228 deliver_drop_privilege is set.
6230 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6231 it now includes up to 150 characters of the unexpected data in the log
6234 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6235 and building data strings. The size of both of these buffers was 10 000
6236 bytes - far larger than anybody would *ever* want, thought I. Needless to
6237 say, somebody hit the limit. I have increased the maximum line length to
6238 20 000 and the maximum data length of concatenated lines to 100 000. I have
6239 also fixed two bugs, because there was no checking on these buffers. Tsk,
6240 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6241 buffer is too small.
6243 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6244 lsearch lookups. Now it does.
6246 46. When parsing a route_list item in a manualroute router, a fixed-length
6247 buffer was used for the list of hosts. I made this 1024 bytes long,
6248 thinking that nobody would ever have a list of hosts that long. Wrong.
6249 Somebody had a whole pile of complicated expansion conditions, and the
6250 string was silently truncated, leading to an expansion error. It turns out
6251 that it is easier to change to an unlimited length (owing to other changes
6252 that have happened since this code was originally written) than to build
6253 structure for giving a limitation error. The length of the item that
6254 expands into the list of hosts is now unlimited.
6256 47. The lsearch lookup could not handle data where the length of text line was
6257 more than 4095 characters. Such lines were truncated, leading to shortened
6258 data being returned. It should now handle lines of any length.
6260 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6261 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6264 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6265 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6266 no longer happens in comment lines. A list of replacements is now placed
6267 at the head of all of the source files, except those whose only change is
6268 to replace PERL_COMMAND in the very first #! line.
6270 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6271 messages on the queue, with a bottom-up merge sort, using code contributed
6272 by Michael Haardt. This should make operations like -bp somewhat faster on
6273 large queues. It won't affect queue runners, except when queue_run_in_order
6276 51. Installed eximstats 1.31 in the distribution.
6278 52. Added support for SRV lookups to the dnslookup router.
6280 53. If an ACL referred to $message_body or $message_body_end, the value was not
6281 reset for any messages that followed in the same SMTP session.
6283 54. The store-handling optimization for building very long strings was not
6284 differentiating between the different store pools. I don't think this
6285 actually made any difference in practice, but I've tidied it.
6287 55. While running the routers to verify a sender address, $sender_address
6288 was still set to the sender address. This is wrong, because when routing to
6289 send a bounce to the sender, it would be empty. Therefore, I have changed
6290 it so that, while verifying a sender address, $sender_address is set to <>.
6291 (There is no change to what happens when verifying a recipient address.)
6293 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6294 target A or AAAA records (if not already returned) without resetting the
6295 qualify_single or search_parents options of the DNS resolver. These are
6296 inappropriate in this case because the targets of MX and SRV records must
6297 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6298 target that, when qualified, matched something in the local domain. These
6299 two options are now turned off when doing these lookups.
6301 57. It seems that at least some releases of Reiserfs (which does not have the
6302 concept of a fixed number of inodes) returns zero and not -1 for the
6303 number of available inodes. This interacted badly with check_spool_inodes,
6304 which assumed that -1 was the "no such thing" setting. What I have done is
6305 to check that the total number of inodes is greater than zero before doing
6306 the test of how many are available.
6308 58. When a "warn" ACL statement has a log_message modifier, the message is
6309 remembered, and not repeated. This is to avoid a lot of repetition when a
6310 message has many recipients that cause the same warning to be written.
6311 However, Exim was preserving the list of already written lines for an
6312 entire SMTP session, which doesn't seem right. The memory is now reset if a
6313 new message is started.
6315 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6316 debugging output unless log_rewrite was also set.
6318 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6319 of (int)(handle) when we know that handle contains (void *)(-1).
6321 61. The Exim daemon panic-logs an error return when it closes the incoming
6322 connection. However "connection reset by peer" seems to be common, and
6323 isn't really an error worthy of noting specially, so that particular error
6326 62. When Exim is trying to find all the local interfaces, it used to panic and
6327 die if the ioctl to get the interface flags failed. However, it seems that
6328 on at least one OS (Solaris 9) it is possible to have an interface that is
6329 included in the list of interfaces, but for which you get a failure error
6330 for this call. This happens when the interface is not "plumbed" into a
6331 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6332 failure of the "get flags" call assumes that the interface is down.
6334 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6335 makes life easier for people who are doing arithmetic on fields extracted
6336 from dates, where you often get leading zeros that should not be
6337 interpreted as octal.
6339 64. Added qualify_domain to the redirect router, to override the global
6342 65. If a pathologically long header line contained very many addresses (the
6343 report of this problem mentioned 10 000) and each of them was rewritten,
6344 Exim could use up a very large amount of memory. (It kept on making new
6345 copies of the header line as it rewrote, and never released the old ones.)
6346 At the expense of a bit more processing, the header rewriting function has
6347 been changed so that it no longer eats memory in this way.
6349 66. The generation of the Received: header has been moved from the time that a
6350 message starts to be received, to the time that it finishes. The timestamp
6351 in the Received: header should now be very close to that of the <= log
6352 line. There are two side-effects of this change:
6354 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6355 logged header lines no longer include the local Received: line, because
6356 it has not yet been created. The same applies to a copy of the message
6357 that is returned to a non-SMTP sender when a message is rejected.
6359 (b) When a filter file is tested using -bf, no additional Received: header
6360 is added to the test message. After some thought, I decided that this
6363 This change does not affect the value of $received_for. It is still set
6364 after address rewriting, but before local_scan() is called.
6366 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6368 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6369 gave an unhelpful panic error message, and a defer error. I have managed to
6370 change this behaviour so that it now rejects any supplied certificate,
6371 which seems right, as the list of acceptable certificates is empty.
6373 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6374 gave an unhelpful defer error. I have not managed to make this reject any
6375 supplied certificates, but the error message it gives is "no certificate
6376 supplied", which is not helpful.
6378 70. exigrep's output now also includes lines that are not associated with any
6379 message, but which match the given pattern. Implemented by a patch from
6380 Martin Sluka, which also tidied up the Perl a bit.
6382 71. Recipient callout verification, like sender verification, was using <> in
6383 the MAIL FROM command. This isn't really the right thing, since the actual
6384 sender may affect whether the remote host accepts the recipient or not. I
6385 have changed it to use the actual sender in the callout; this means that
6386 the cache record is now keyed on a recipient/sender pair, not just the
6387 recipient address. There doesn't seem to be a real danger of callout loops,
6388 since a callout by the remote host to check the sender would use <>.
6389 [SEE ABOVE: changed after hitting problems.]
6391 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6392 temporary errors. However, in the case of such a code being given after
6393 the end of a data transmission (i.e. after ".") Exim was failing to write
6394 a retry record for the message. (Yes, there was some broken host that was
6395 actually sending 8xx at this point.)
6397 73. An unknown lookup type in a host list could cause Exim to panic-die when
6398 the list was checked. (An example that provoked this was putting <; in the
6399 middle of a list instead of at the start.) If this happened during a DATA
6400 ACL check, a -D file could be left lying around. This kind of configuration
6401 error no longer causes Exim to die; instead it causes a defer error. The
6402 incident is still logged to the main and panic logs.
6404 74. Buglet left over from Exim 3 conversion. The message "too many messages
6405 in one connection" was written to the rejectlog but not the mainlog, except
6406 when address rewriting (yes!) was being logged.
6408 75. Added write_rejectlog option.
6410 76. When a system filter was run not as root (that is, when system_filter_user
6411 was set), the values of the $n variables were not being returned to the
6412 main process; thus, they were not subsequently available in the $sn
6415 77. Added +return_path_on_delivery log selector.
6417 78. A connection timeout was being treated differently from recipients deferred
6418 when testing hosts_max_try with a message that was older than the host's
6419 retry timeout. (The host should not be counted, thus allowing all hosts to
6420 be tried at least once before bouncing.) This may have been the cause of an
6421 occasionally reported bug whereby a message would remain on the queue
6422 longer than the retry timeout, but would be bounced if a delivery was
6423 forced. I say "may" because I never totally pinned down the problem;
6424 setting up timeout/retry tests is difficult. See also the next item.
6426 79. The ultimate address timeout was not being applied to errors that involved
6427 a combination of host plus message (for example, a timeout on a MAIL
6428 command). When an address resolved to a number of possible hosts, and they
6429 were not all tried for each delivery (e.g. because of hosts_max_try), a
6430 message could remain on the queue longer than the retry timeout.
6432 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6435 81. Fixed an obscure SMTP outgoing bug which required at least the following
6436 conditions: (a) there was another message waiting for the same server;
6437 (b) the server returned 5xx to all RCPT commands in the first message so
6438 that the message was not completed; (c) the server dropped the connection
6439 or gave a negative response to the RSET that Exim sends to abort the
6440 transaction. The observed case was a dropped connection after DATA that had
6441 been sent in pipelining mode. That is, the server had advertised PIPELINING
6442 but was not implementing it correctly. The effect of the bug was incorrect
6443 behaviour, such as trying another host, and this could lead to a crash.
6449 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6450 and daemon.c were passed as pointers to ints; they should have been
6451 pointers to socklen_t variables (which are typically unsigned ints).
6453 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6456 3. Fixed a really odd bug that affected only the testing scheme; patching a
6457 certain fixed string in the binary changed the value of another string that
6458 happened to be identical to the end of the original first string.
6460 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6461 name", it returns that address as the IP address. On some operating
6462 systems (e.g. Solaris), it also passes back the IP address string as the
6463 "host name". However, on others (e.g. Linux), it passes back an empty
6464 string. Exim wasn't checking for this, and was changing the host name to an
6465 empty string, assuming it had been canonicalized.
6467 5. Although rare, it is permitted to have more than one PTR record for a given
6468 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6469 all the names associated with an address, because they do in Solaris.
6470 However, it seems that they do not in Linux for data that comes from the
6471 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6472 I found this out when I moved to a new Linux workstation and tried to run
6473 the Exim test suite.
6475 To get round this problem I have changed the code so that it now does its
6476 own call to the DNS to look up PTR records when searching for a host name.
6477 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6478 addresses that are only in /etc/hosts are still found.
6480 This behaviour is, however, controlled by an option called host_lookup_
6481 order, which defaults to "bydns:byaddr". If people want to use the other
6482 order, or indeed, just use one or the other means of lookup, they can
6483 specify it in this variable.
6485 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6486 some operating systems, this comes back from gethostbyaddr() as an empty
6487 string, and this is what Exim used to test for. However, it seems that in
6488 other systems, "." is yielded. Exim now tests for this case too.
6490 7. The values of check_spool_space and check_log_space are now held internally
6491 as a number of kilobytes instead of an absolute number of bytes. If a
6492 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6493 kilobyte. This means that much larger values can be stored.
6495 8. Exim monitor: an attempt to get the action menu when not actually pointing
6496 at a message produces an empty menu entitled "No message selected". This
6497 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6498 no entries in it ("Shell widget menu has zero width and/or height"). So I
6499 have added a single, blank menu entry in this case.
6501 9. Added ${quote_local_part.
6503 10. MIME decoding is now applied to the contents of Subject: header lines when
6506 11. Now that a reference to $sender_host_address automatically causes a reverse
6507 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6508 host lookup before query-style lookups in lists that might use this
6509 variable. This has therefore been abolished, and the "net-" prefix is no
6510 longer necessary for query-style lookups.
6512 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6513 have been a typo for LFLAGS, so it has been changed.
6515 13. The install script calls Exim with "-C /dev/null" in order to find the
6516 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6517 to be output. However, since Exim outputs its version number before the
6518 error, it didn't break the script. It just looked ugly. I fixed this by
6519 always allowing "-C /dev/null" if the caller is root.
6521 14. Ignore overlarge ACL variable number when reading spool file - insurance
6522 against a later release with more variables having written the file.
6524 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6525 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6526 this, as well as the form without the "IPv6" on the front.
6528 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6529 OS/Makefile-Darwin file.
6531 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6532 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6534 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6535 variable. However, some OS use ints for inodes. I've added cast to long int
6536 to get rid of the compiler warning.
6538 19. I had forgotten to lock out "/../" in configuration file names when
6539 ALT_CONFIG_PREFIX was set.
6541 20. Routers used for verification do not need to specify transports. However,
6542 if such a router generated a host list, and callout was configured, Exim
6543 crashed, because it could not find a port number from the (non-existent)
6544 transport. It now assumes port 25 in this circumstance.
6546 21. Added the -t option to exigrep.
6548 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6549 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6550 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6551 them is set, LOOKUP_LSEARCH is forced.
6553 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6555 24. Added sender and host information to the "rejected by local_scan()" log
6556 line; previously there was no indication of these.
6558 25. Added .include_if_exists.
6560 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6561 Linux. It turns out that not all file systems support this. Apparently some
6562 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6563 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6564 not supported on the file descriptor, is now ignored when Exim is trying to
6565 sync a directory. This applies only to Linux.
6567 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6569 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6570 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6573 29. Added support for maildirsize files from supplied patch (modified a bit).
6575 30. The use of :fail: followed by an empty string could lead Exim to respond to
6576 sender verification failures with (e.g.):
6578 550 Verification failed for <xxx>
6579 550 Sender verify failed
6581 where the first response line was missing the '-' that indicates it is not
6582 the final line of the response.
6584 31. The loop for finding the name of the user that called Exim had a hardwired
6585 limit of 10; it now uses the value of finduser_retries, which is used for
6586 all other user lookups.
6588 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6591 33. Exim was neglecting to zero errno before one call of strtol() when
6592 expanding a string and expecting an integer value. On some systems this
6593 resulted in spurious "integer overflow" errors. Also, it was casting the
6594 result into an int without checking.
6596 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6597 did not work. The code looks as if it has *never* worked, though it appears
6598 to have been documented since at least release 1.62. I have made it work.
6600 35. The "timeout_DNS" error in retry rules, also documented since at least
6601 1.62, also never worked. As it isn't clear exactly what this means, and
6602 clearly it isn't a major issue, I have abolished the feature by treating it
6603 as "timeout", and writing a warning to the main and panic logs.
6605 36. The display of retry rules for -brt wasn't always showing the error code
6608 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6609 timeout_connect_A, timeout_connect_MX.
6611 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6612 to the empty sender.
6614 39. The daemon was not analysing the content of -oX till after it had closed
6615 stderr and disconnected from the controlling terminal. This meant that any
6616 syntax errors were only noted on the panic log, and the return code from
6617 the command was 0. By re-arranging the code a little, I've made the
6618 decoding happen first, so such errors now appear on stderr, and the return
6619 code is 1. However, the actual setting up of the sockets still happens in
6620 the disconnected process, so errors there are still only recorded on the
6623 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6624 connections (as happens on some IP stacks) was logged at start up time as
6625 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6626 it from "IPv6 and IPv4", which means that two separate sockets are being
6629 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6630 says whether AF_INET or AF_INET6 was passed as an argument.
6632 42. Exiwhat output was messed up when time zones were included in log
6635 43. Exiwhat now gives more information about the daemon's listening ports,
6636 and whether -tls-on-connect was used.
6638 44. The "port" option of the smtp transport is now expanded.
6640 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6641 silently ignored. Now an error message is written to the main and panic
6644 46. There's a new ACL modifier called "logwrite" which writes to a log file
6645 as soon as it is encountered.
6647 47. Added $local_user_uid and $local_user_gid at routing time.
6649 48. Exim crashed when trying to verify a sender address that was being
6652 49. Exim was recognizing only a space character after ".include". It now also
6653 recognizes a tab character.
6655 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6656 extracting the relevant information from the specification. The man page no
6657 longer contains scrambled data for the -d option, and I've added a section
6658 at the front about calling Exim under different names.
6660 51. Added "extra_headers" argument to the "mail" command in filter files.
6662 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6665 53. Installed eximstats 1.29.
6667 54. Added transport_filter_timeout as a generic transport option.
6669 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6670 Cc: header lines. This was required by RFC 822, but it not required by RFC
6673 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6674 incoming messages that did not have them. Now it does so only if the
6675 message originates locally, that is, if there is no associated remote host
6676 address. When Resent- header lines are present, this applies to the Resent-
6677 lines rather than the non-Resent- lines.
6679 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6680 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6682 58. Messages for configuration errors now include the name of the main
6683 configuration file - useful now that there may be more than one file in a
6684 list (.included file names were always shown).
6686 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6687 for those rare installations that do not start the daemon as root or run it
6688 setuid root. I've cut out the call to initgroups() if the daemon is not
6691 60. The Exim user and group can now be bound into the binary as text strings
6692 that are looked up at the start of Exim's processing.
6694 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6696 62. Added $mailstore_basename variable.
6698 63. Installed patch to sieve.c from Michael Haardt.
6700 64. When Exim failed to open the panic log after failing to open the main log,
6701 the original message it was trying to log was written to stderr and debug
6702 output, but if they were not available (the usual case in production), it
6703 was lost. Now it is written to syslog before the two lines that record the
6704 failures to open the logs.
6706 65. Users' Exim filters run in subprocesses under the user's uid. It is
6707 possible for a "deliver" command or an alias in a "personal" command to
6708 provoke an address rewrite. If logging of address rewriting is configured,
6709 this fails because the process is not running as root or exim. There may be
6710 a better way of dealing with this, but for the moment (because 4.30 needs
6711 to be released), I have disabled address rewrite logging when running a
6712 filter in a non-root, non-exim process.
6718 1. The buildconfig auxiliary program wasn't quoting the value set for
6719 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6720 not defined. This bug was masked in 4.22 by the effect that was fixed in
6723 2. Some messages that were rejected after a message id was allocated were
6724 shown as "incomplete" by exigrep. It no longer does this for messages that
6725 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6727 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6728 have allow_domain_literals set, the ID did not get logged in the <= line.
6729 Domain literals are now always recognized in Message-ID: header lines.
6731 4. The first argument for a ${extract expansion item is the key name or field
6732 number. Leading and trailing spaces in this item were not being ignored,
6733 causing some misleading effects.
6735 5. When deliver_drop_privilege was set, single queue runner processes started
6736 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6737 same command in the process it spins off) were not dropping privilege.
6739 6. When the daemon running as "exim" started a queue runner, it always
6740 re-executed Exim in the spun-off process. This is a waste of effort when
6741 deliver_drop_privilege is set. The new process now just calls the
6742 queue-runner function directly.
6748 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6751 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6752 ignored. Though the use of -oP was forcing the writing of a pid file, it
6753 was always written to the default place.
6755 3. If the message "no IP address found for host xxxx" is generated during
6756 incoming verification, it is now followed by identification of the incoming
6757 connection (so you can more easily find what provoked it).
6759 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6761 5. Added some features to "harden" Exim a bit more against certain attacks:
6763 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6764 be put in Local/Makefile. This is like the never_users runtime option,
6765 but it cannot be overridden. The default setting is "root".
6767 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6768 prefix string with which any file named in a -C command line option
6771 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6772 is retained for -C and -D only if the caller of Exim is root. Without
6773 it, the exim user may also use -C and -D and retain privilege.
6775 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6776 command line option is disabled.
6778 6. Macro names set by the -D option must start with an upper case letter, just
6779 like macro names defined in the configuration file.
6781 7. Added "dereference=" facility to LDAP.
6783 8. Two instances of the typo "uknown" in the source files are fixed.
6785 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6786 the Configure-Makefile script screwed up while processing it.
6788 10. Incorporated PCRE 4.4.
6790 11. The SMTP synchronization check was not operating right at the start of an
6791 SMTP session. For example, it could not catch a HELO sent before the client
6792 waited for the greeting. There is now a check for outstanding input at the
6793 point when the greeting is written. Because of the duplex, asynchronous
6794 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6795 way, but not yet received, when the check is performed.
6797 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6798 on TCP/IP sockets, because this apparently causes some broken clients to
6801 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6802 unchanged) from the Cygwin maintainer.
6804 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6805 was supported instead of testing for mbx. Effectively a typo.
6807 15. The spa authenticator server code was not checking that the input it
6808 received was valid base64.
6810 16. The debug output line for the "set" modifier in ACLs was not showing the
6811 name of the variable that was being set.
6813 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6815 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6816 lookup on its own. Something else was needed to trigger the lookup. For
6817 example, a match in host_lookup or the need for a host name in a host list.
6818 Now, if $sender_host_name is referenced and the host name has not yet been
6819 looked up, a lookup is performed. If the lookup fails, the variable remains
6820 empty, and $host_lookup_failed is set to "1".
6822 19. Added "eqi" as a case-independent comparison operator.
6824 20. The saslauthd authentication condition could segfault if neither service
6825 nor realm was specified.
6827 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6828 error message that was logged was misleading, and incoming SMTP
6829 connections were dropped. The message is now more accurate, and temporary
6830 errors are given to SMTP connections.
6832 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6833 (except RSET) with a 421 error, until QUIT is received. However, it was
6834 failing to send a response to QUIT.
6836 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6837 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6838 ACL is now run after the helo_try_verify_hosts code.
6840 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6841 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6842 case-independent, so other case variants are also recognized). Apparently
6843 some systems use these upper case variants.
6845 25. If more than two messages were waiting for the same host, and a transport
6846 filter was specified for the transport, Exim sent two messages over the
6847 same TCP/IP connection, and then failed with "socket operation on non-
6848 socket" when it tried to send the third.
6850 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6852 27. The extern definition of crypt16() in expand.c was not being excluded when
6853 the OS had its own crypt16() function.
6855 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6856 as a preferred synonym for return_size_limit, both as an option and as an
6859 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6861 30. Changed the default configuration ACL to relax the local part checking rule
6862 for addresses that are not in any local domains. For these addresses,
6863 slashes and pipe symbols are allowed within local parts, but the sequence
6864 /../ is explicitly forbidden.
6866 31. SPA server authentication was not clearing the challenge buffer before
6869 32. log_message in a "warn" ACL statement was writing to the reject log as
6870 well as to the main log, which contradicts the documentation and doesn't
6871 seem right (because no rejection is happening). So I have stopped it.
6873 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6874 However, I am unable to do any testing of this.
6876 34. Fixed an infelicity in the appendfile transport. When checking directories
6877 for a mailbox, to see if any needed to be created, it was accidentally
6878 using path names with one or more superfluous leading slashes; tracing
6879 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6881 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6882 contents are added to the log line that is written for every discarded
6883 recipient. (Previously a log_message setting was ignored.)
6885 36. The ${quote: operator now quotes the string if it is empty.
6887 37. The install script runs exim in order to find its version number. If for
6888 some reason other than non-existence or emptiness, which it checks, it
6889 could not run './exim', it was installing it with an empty version number,
6890 i.e. as "exim-". This error state is now caught, and the installation is
6893 38. An argument was missing from the function that creates an error message
6894 when Exim fails to connect to the socket for saslauthd authentication.
6895 This could cause Exim to crash, or give a corrupted message.
6897 39. Added isip, isip4, and isip6 to ${if conditions.
6899 40. The ACL variables $acl_xx are now saved with the message, and can be
6900 accessed later in routers, transports, and filters.
6902 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6903 strings in the file are not string-expanded.
6905 42. If a MAIL command specified a SIZE value that was too large to fit into an
6906 int variable, the check against message_size_limit failed. Such values are
6907 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6908 day this will have to be increased, but I don't think I want to be around
6909 when emails are that large.
6916 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6917 iconv() is not standard in FreeBSD.
6919 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6920 IPv6 enabled. The observed symptom was a segmentation fault on return from
6921 the function os_common_find_running_interfaces() in src/os.c.
6923 3. In the check_special_case() function in daemon.c I had used "errno" as an
6924 argument name, which causes warnings on some systems. This was basically a
6925 typo, since it was named "eno" in the comments!
6927 4. The code that waits for the clock to tick (at a resolution of some fraction
6928 of a second) so as to ensure message-id uniqueness was always waiting for
6929 at least one whole tick, when it could have waited for less. [This is
6930 almost certainly not relevant at current processor speeds, where it is
6931 unlikely to ever wait at all. But we try to future-proof.]
6933 5. The function that sleeps for a time interval that includes fractions of a
6934 second contained a race. It did not block SIGALRM between setting the
6935 timer, and suspending (a couple of lines later). If the interval was short
6936 and the sigsuspend() was delayed until after it had expired, the suspension
6937 never ended. On busy systems this could lead to processes getting stuck for
6940 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6941 process, before it forks any delivery processes. The open lookup caching
6942 mechanism meant that the open file or database connection was passed into
6943 the delivery process. The problem was that delivery processes always tidy
6944 up cached lookup data. This could cause a problem for the next delivery
6945 process started by the queue runner, because the external queue runner
6946 process does not know about the closure. So the next delivery process
6947 still has data in the lookup cache. In the case of a file lookup, there was
6948 no problem because closing a file descriptor in a subprocess doesn't affect
6949 the parent. However, if the lookup was caching a connection to a database,
6950 the connection was closed, and the second delivery process was likely to
6951 see errors such as "PGSQL: query failed: server closed the connection
6952 unexpectedly". The problem has been fixed by closing all cached lookups
6953 in a queue runner before running a delivery process.
6955 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6956 seem to have the "const" qualifier which it has on other OS. I've
6959 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6960 *of the same type* (client or server) with the same public name that an
6961 error should be diagnosed.
6963 9. When Exim looked up a host name for an IP address, but failed to find the
6964 original IP address when looking up the host name (a safety check), it
6965 output the message "<ip address> does not match any IP for NULL", which was
6966 confusing, to say the least. The bug was that the host name should have
6967 appeared instead of "NULL".
6969 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6970 user that is built into the binary, and the -C or -D options is used, root
6971 privilege is dropped before the configuration file is read. In addition,
6972 logging is switched to stderr instead of the normal log files. If the
6973 configuration then re-defines the Exim user, the unprivileged environment
6974 is probably not what is expected, so Exim logs a panic warning message (but
6977 However, if deliver_drop_privilege is set, the unprivileged state may well
6978 be exactly what is intended, so the warning has been cut out in that case,
6979 and Exim is allowed to try to write to its normal log files.
6985 1. smtp_return_error_details was not giving details for temporary sender
6986 or receiver verification errors.
6988 2. Diagnose a configuration error if two authenticators have the same public
6991 3. Exim used not to create the message log file for a message until the first
6992 delivery attempt. This could be confusing when incoming messages were held
6993 for policy or load reasons. The message log file is now created at the time
6994 the message is received, and an initial "Received" line is written to it.
6996 4. The automatically generated man page for command line options had a minor
6997 bug that caused no ill effects; however, a more serious problem was that
6998 the procedure for building the man page automatically didn't always
6999 operate. Consequently, release 4.20 contains an out-of-date version. This
7000 shouldn't happen again.
7002 5. When building Exim with embedded Perl support, the script that builds the
7003 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7004 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7006 6. The freeze_tell option was not being used for messages that were frozen on
7007 arrival, either by an ACL or by local_scan().
7009 7. Added the smtp_incomplete_transaction log selector.
7011 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7012 was accepting AUTH without a new EHLO.
7014 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7015 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7018 10. Logging of TCP/IP connections (when configured) now happens in the main
7019 daemon process instead of the child process, so that the TCP/IP connection
7020 count is more accurate (but it can never be perfect).
7022 11. The use of "drop" in a nested ACL was not being handled correctly in the
7023 outer ACL. Now, if condition failure induced by the nested "drop" causes
7024 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7025 or "require"), the connection is dropped.
7027 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7028 that yield "discard" can now be used with an "accept" or a "discard" verb,
7029 but an error is generated for any others (because I can't see a useful way
7030 to define what should happen).
7032 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7033 are now processed in the same way as lines in the Exim configuration file.
7034 In particular, continuation lines are supported.
7036 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7038 15. Added -ti meaning -t -i.
7040 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7041 domains, and warn by logging if others are found.
7043 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7044 returned from GIFCONF when Exim is trying to find the list of interfaces on
7045 a host. The code in os.c has been modified to copy each ifreq to an aligned
7046 structure in all cases.
7048 Also, in some cases, the returned ifreq's were being copied to a 'struct
7049 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7050 means the last couple of bytes of an IPv6 address could be chopped if the
7051 ifreq contained only a normal sockaddr (14 bytes storage).
7053 18. Named domain lists were not supported in the hosts_treat_as_local option.
7054 An entry such as +xxxx was not recognized, and was treated as a literal
7057 19. Ensure that header lines added by a DATA ACL are included in the reject log
7058 if the ACL subsequently rejects the message.
7060 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7061 MD5 (which is deprecated).
7063 21. When testing a filter file using -bf, Exim was writing a message when it
7064 took the sender from a "From " line in the message, but it was not doing so
7065 when it took $return_path from a Return-Path: header line. It now does.
7067 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7068 with a valid header line field name (a series of printing characters
7069 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7071 23. Changed "disc" in the source to "disk" to conform to the documentation and
7072 the book and for uniformity.
7074 24. Ignore Sendmail's -Ooption=value command line item.
7076 25. When execve() failed while trying to run a command in a pipe transport,
7077 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7078 could be confused with a return value of 69 from the command itself. This
7079 has been changed to 127, the value the shell returns if it is asked to run
7080 a non-existent command. The wording for the related log line suggests a
7081 non-existent command as the problem.
7083 26. If received_header_text expands to an empty string, do not add a Received:
7084 header line to the message. (Well, it adds a token one on the spool, but
7085 marks it "old" so that it doesn't get used or transmitted.)
7087 27. Installed eximstats 1.28 (addition of -nt option).
7089 28. There was no check for failure on the call to getsockname() in the daemon
7090 code. This can fail if there is a shortage of resources on the system, with
7091 ENOMEM, for example. A temporary error is now given on failure.
7093 29. Contrary to the C standard, it seems that in some environments, the
7094 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7095 program. Exim now does this explicitly; it affects the formatting of
7096 timestamps using strftime().
7098 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7099 complaints. I've now initialized all the variables, to avoid this.
7101 32. Header lines added by a system filter were not being "seen" during
7102 transport-time rewrites.
7104 33. The info_callback() function passed to OpenSSL is set up with type void
7105 (*)(SSL *, int, int), as described somewhere. However, when calling the
7106 function (actually a macro) that sets it up, the type void(*)() is
7107 expected. I've put in a cast to prevent warnings from picky compilers.
7109 34. If a DNS black list lookup found a CNAME record, but there were no A
7110 records associated with the domain it pointed at, Exim crashed.
7112 35. If a DNS black list lookup returned more than one A record, Exim ignored
7113 all but the first. It now scans all returned addresses if a particular IP
7114 value is being sought. In this situation, the contents of the
7115 $dnslist_value variable are a list of all the addresses, separated by a
7118 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7119 to accept a host name and all its aliases if the forward lookup for any of
7120 them yielded the IP address of the incoming connection. Now it accepts only
7121 those names whose forward lookup yields the correct IP address. Any other
7122 names are discarded. This closes a loophole whereby a rogue DNS
7123 administrator could create reverse DNS records to break through a
7124 wildcarded host restriction in an ACL.
7126 37. If a user filter or a system filter that ran in a subprocess used any of
7127 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7128 the wrong values were passed to the pipe command ($thisaddress had the
7129 value of $0, $0 had the value of $1, etc). This bug was introduced by
7130 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7132 38. Improved the line breaking for long SMTP error messages from ACLs.
7133 Previously, if there was no break point between 40 and 75 characters, Exim
7134 left the rest of the message alone. Two changes have been made: (a) I've
7135 reduced the minimum length to 35 characters; (b) if it can't find a break
7136 point between 35 and 75 characters, it looks ahead and uses the first one
7137 that it finds. This may give the occasional overlong line, but at least the
7138 remaining text gets split now.
7140 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7141 file descriptors might be low, and that setting 1000 would always raise it.
7142 It turns out that in some environments, the limit is already over 1000 and
7143 that lowering it causes trouble. So now Exim takes care not to decrease it.
7145 40. When delivering a message, the value of $return_path is set to $sender_
7146 address at the start of routing (routers may change the value). By an
7147 oversight, this default was not being set up when an address was tested by
7148 -bt or -bv, which affected the outcome if any router or filter referred to
7151 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7152 log only when "message" or "log_message" are set. However, if one of the
7153 conditions was an address verification, or a call to a nested ACL, the
7154 messages generated by the underlying test were being passed through. This
7155 no longer happens. The underlying message is available in $acl_verify_
7156 message for both "message" and "log_message" expansions, so it can be
7157 passed through if needed.
7159 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7160 new expansion $bh_ to give the encoded byte string without charset
7161 translation. Translation happens only if iconv() is available; HAVE_ICONV
7162 indicates this at build time. HEADERS_CHARSET gives the charset to
7163 translate to; headers_charset can change it in the configuration, and
7164 "headers charset" can change it in an individual filter file.
7166 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7167 that creates RFC 2047 encoded "words" labels them as that charset instead
7168 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7169 expansion operator; (ii) when Exim creates a From: line for a local
7170 message; (iii) when a header line is rewritten to include a "phrase" part.
7172 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7173 buggy, causing it to skip the first lines of messages whose message ID
7174 ended in 'D'. This would not have bitten before Exim release 4.14, because
7175 message IDs were unlikely to end in 'D' before then. The effect was to have
7176 incorrect size information for certain domains.
7178 45. #include "config.h" was missing at the start of the crypt16.c module. This
7179 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7182 46. If there was a timeout during a "random" callout check, Exim treated it as
7183 a failure of the random address, and carried on sending RSET and the real
7184 address. If the delay was just some slowness somewhere, the response to the
7185 original RCPT would be taken as a response to RSET and so on, causing
7186 mayhem of various kinds.
7188 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7189 when I implemented it. It didn't allow for the fact that some option values
7190 may legitimately be negative (e.g. size_addition), and it didn't even do
7191 the right test for positive values.
7193 48. Domain names in DNS records are case-independent. Exim always looks them up
7194 in lower case. Some resolvers return domain names in exactly the case they
7195 appear in the zone file, that is, they may contain uppercase letters. Not
7196 all resolvers do this - some return always lower case. Exim was treating a
7197 change of case by a resolver as a change of domain, similar to a widening
7198 of a domain abbreviation. This triggered its re-routing code and so it was
7199 trying to route what was effectively the same domain again. This normally
7200 caused routing to fail (because the router wouldn't handle the domain
7201 twice). Now Exim checks for this case specially, and just changes the
7202 casing of the domain that it ultimately uses when it transmits the message
7205 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7208 50. If a filter generated a file delivery with a non-absolute name (possible if
7209 no home directory exists for the router), the forbid_file option was not
7212 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7213 the existing equality matching.
7215 52. Exim was using ints instead of ino_t variables in some places where it was
7216 dealing with inode numbers.
7218 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7219 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7220 called TMPDIR, and if it finds it is different, it changes its value.
7222 54. The smtp_printf() function is now made available to local_scan() so
7223 additional output lines can be written before returning. There is also an
7224 smtp_fflush() function to enable the detection of a dropped connection.
7225 The variables smtp_input and smtp_batched_input are exported to
7228 55. Changed the default runtime configuration: the message "Unknown user"
7229 has been removed from the ACL, and instead placed on the localuser router,
7230 using the cannot_route_message feature. This means that any verification
7231 failures that generate their own messages won't get overridden. Similarly,
7232 the "Unrouteable address" message that was in the ACL for unverifiable
7233 relay addresses has also been removed.
7235 56. Added hosts_avoid_esmtp to the smtp transport.
7237 57. The exicyclog script was not checking for the esoteric option
7238 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7239 will work only if exicyclog is run under the appropriate euid.
7241 58. Following a discussion on the list, the rules by which Exim recognises line
7242 endings on incoming messages have been changed. The -dropcr and drop_cr
7243 options are now no-ops, retained only for backwards compatibility. The
7244 following line terminators are recognized: LF CRLF CR. However, special
7245 processing applies to CR:
7247 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7248 nor a local message in the state where . is a terminator.
7250 (ii) If a bare CR is encountered in a header line, an extra space is added
7251 after the line terminator so as not to end the header. The reasoning
7252 behind this is that bare CRs in header lines are most likely either
7253 to be mistakes, or people trying to play silly games.
7255 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7256 was being incorrectly given as 18 bytes larger than it should have been.
7257 This is a VOB (very old bug).
7259 60. This may never have affected anything current, but just in case it has:
7260 When the local host is found other than at the start of a list of hosts,
7261 the local host, those with the same MX, and any that follow, are discarded.
7262 When the list in question was part of a longer list of hosts, the following
7263 hosts (not currently being processed) were also being discarded. This no
7264 longer happens. I'm not sure if this situation could ever has previously
7267 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7270 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7271 header. This is something that is recommended in a new Internet Draft, and
7272 is something that is documented as being done by Sendmail. There are two
7273 possible values. For messages generated by the autoreply transport, Exim
7276 Auto-Submitted: auto-replied
7278 whereas for all other generated messages (e.g. bounces) it adds
7280 Auto-Submitted: auto-generated
7282 63. The "personal" condition in filters now includes a test for the
7283 Auto-Submitted: header. If it contains the string "auto-" the message it
7284 not considered personal.
7286 64. Added rcpt_include_affixes as a generic transport option.
7288 65. Added queue_only_override (default true).
7290 66. Added the syslog_duplication option.
7292 67. If what should have been the first header line of a message consisted of
7293 a space followed by a colon, Exim was mis-interpreting it as a header line.
7294 It isn't of course - it is syntactically invalid and should therefore be
7295 treated as the start of the message body. The misbehaviour could have
7296 caused a number of strange effects, including loss of data in subsequent
7297 header lines, and spool format errors.
7299 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7300 client host had authenticated. This control can now be exercised by an ACL
7301 for more flexibility.
7303 69. By default, callouts do not happen when testing with -bh. There is now a
7304 variant, -bhc, which does actually run the callout code, including
7305 consulting and updating the callout cache.
7307 70. Added support for saslauthd authentication, courtesy of Alexander
7310 71. If statvfs() failed on the spool or log directories while checking their
7311 size for availability, Exim confusingly gave the error "space shortage".
7312 Furthermore, in debugging mode it crashed with a floating point exception.
7313 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7314 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7315 serious problem, Exim now writes to the main and panic logs when this
7316 happens, with details of the failure. It then refuses to accept the
7317 incoming message, giving the message "spool directory problem" or "log
7318 directory problem" with a 421 code for SMTP messages.
7320 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7321 0, 1, and 2 exist, because some OS complain for execs without them (see
7322 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7323 descriptors. However, the code omitted to check that the open succeeded,
7324 causing mysterious errors if for some reason the permissions on /dev/null
7325 got screwed. Now Exim writes a message to the main and panic logs, and
7326 bombs out if it can't open /dev/null.
7328 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7329 interact so that it is all more flexible. It is supposed to remain
7330 backwards compatible. Also added extra_local_interfaces.
7332 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7333 to bomb out with an assertion failure - to the client this appears as a
7334 connection drop. This problem occurs in the part of the code that was taken
7335 from the Samba project. Fortunately, the assertion is in a very simple
7336 function, so I have fixed this by reproducing the function inline in the
7337 one place where it is called, and arranging for authentication to fail
7338 instead of killing the process with assert().
7340 75. The SPA client code was not working when the server requested OEM rather
7341 than Unicode encoding.
7343 76. Added code to make require_files with a specific uid setting more usable in
7344 the case where statting the file as root fails - usually a non-root-mounted
7345 NFS file system. When this happens and the failure is EACCES, Exim now
7346 forks a subprocess and does the per-uid checking as the relevant uid.
7348 77. Added process_log_path.
7350 78. If log_file_path was not explicitly set, a setting of check_log_space or
7351 check_log_inodes was ignored.
7353 79. If a space check for the spool or log partitions fails, the incident is now
7354 logged. Of course, in the latter case the data may get lost...
7356 80. Added the %p formatting code to string_format() so that it can be used to
7357 print addresses in debug_print(). Adjusted all the address printing in the
7358 debugging in store.c to use %p rather than %d.
7360 81. There was a concern that a line of code in smtp_in.c could overflow a
7361 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7362 initially expressed, the concern was not well-founded, because trailing
7363 spaces are removed early. However, if the trailing spaces were followed by
7364 a NULL, they did not get removed, so the overflow was possible. Two fixes
7367 (a) I re-wrote the offending code in a cleaner fashion.
7368 (b) If an incoming SMTP command contains a NULL character, it is rejected
7371 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7372 runs initgroups(), so that if the Exim user is in any additional groups,
7373 they will be used during message reception.
7379 The change log for 4.20 and earlier releases has been archived.