1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually removed using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
119 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
120 a "Gnu special" function, asprintf() in the DB utility binary builds; I
121 hope that is portable enough.
123 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
124 requiring a known-CA anchor certificate; make it now rely entirely on the
125 TLSA as an anchor. Checking the name on the leaf cert against the name
126 on the A-record for the host is still done for TA (but not for EE mode).
128 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
129 included in delivery lines for non-proxied connections, when compiled with
130 SUPPORT_SOCKS and running with proxy logging enabled.
132 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
133 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
134 move the existing event to fire before the normal logging of message
135 failure so that custom logging is bracketed by normal logging.
137 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
138 msg:fail:internal event. Developement by Matthias Kurz.
140 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
141 far too small for todays use of crypto signatures stored there. Go all
142 the way to the max DNS message size of 64kB, even though this might be
143 overmuch for IOT constrained device use.
145 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
146 copy a string over itself. The library routine is documented as not
147 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
149 JH/32 For main options check_spool_space and check_inode_space, where the
150 platform supports 64b integers, support more than the previous 2^31 kB
151 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
152 the previous G, M, k.
154 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
155 $authenticated_fail_id variable on authentication failure. Previously
158 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
159 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
160 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
161 GNUTLS_SEC_PARAM_MEDIUM.
163 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
164 side. Previously we would continue as if no SNI had been received.
166 JH/36 Harden the handling of string-lists. When a list consisted of a sole
167 "<" character, which should be a list-separator specification, we walked
168 off past the nul-terimation.
170 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
171 causes) even when the retry time is not yet met. Previously they were
172 not, meaning that when (say) an account was over-quota and temp-rejecting,
173 and multiple senders' messages were queued, only one sender would get
174 notified on each configured delay_warning cycle.
176 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
178 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
179 cipher-suites, an error can be left on the stack even for a succeeding
180 accept; this results in impossible error messages when a later operation
183 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
184 return error codes indicating retry. Under TLS1.3 this becomes required.
186 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
187 it only wrote the new authenticators, resulting in a lack of tracking of
188 peer changes of ESMTP extensions until the next cache flush.
190 JH/41 Fix the loop reading a message header line to check for integer overflow,
191 and more-often against header_maxsize. Previously a crafted message could
192 induce a crash of the recive process; now the message is cleanly rejected.
194 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
195 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
201 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
202 When redis_servers is set to a list of > 1 element, and the Redis servers
203 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
204 case of MOVED into a DEFER case instead, thus moving the query onto the
205 next server in the list. For a cluster of N elements, all N servers must
206 be defined in redis_servers.
208 GF/02 Catch and remove uninitialized value warning in exiqsumm
209 Check for existence of @ARGV before looking at $ARGV[0]
211 JH/01 Replace the store_release() internal interface with store_newblock(),
212 which internalises the check required to safely use the old one, plus
213 the allocate and data copy operations duplicated in both (!) of the
214 extant use locations.
216 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
217 modifier. This matches the restriction on the commandline.
219 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
220 Previously only the last row was returned.
222 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
223 we assumed that tags in the header were well-formed, and parsed the
224 element content after inspecting only the first char of the tag.
225 Assumptions at that stage could crash the receive process on malformed
228 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
229 While running the DKIM ACL we operate on the Permanent memory pool so that
230 variables created with "set" persist to the DATA ACL. Also (at any time)
231 DNS lookups that fail create cache records using the Permanent pool. But
232 expansions release any allocations made on the current pool - so a dnsdb
233 lookup expansion done in the DKIM ACL releases the memory used for the
234 DNS negative-cache, and bad things result. Solution is to switch to the
235 Main pool for expansions.
236 While we're in that code, add checks on the DNS cache during store_reset,
237 active in the testsuite.
238 Problem spotted, and debugging aided, by Wolfgang Breyha.
240 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
241 When none of the hosts presented to a transport match an already-open
242 connection, close it and proceed with the list. Previously we would
243 queue the message. Spotted by Lena with Yahoo, probably involving
246 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
247 Previously a spurious "250 OK id=" response was appended to the proper
250 JH/08 The "support for" informational output now, which built with Content
251 Scanning support, has a line for the malware scanner interfaces compiled
252 in. Interface can be individually included or not at build time.
254 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
255 by the template makefile "src/EDITME". The "STREAM" support for an older
256 ClamAV interface method is removed.
258 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
259 rows affected is given instead).
261 JH/11 The runtime Berkeley DB library version is now additionally output by
262 "exim -d -bV". Previously only the compile-time version was shown.
264 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
265 SMTP connection. Previously, when one had more receipients than the
266 first, an abortive onward connection was made. Move to full support for
267 multiple onward connections in sequence, handling cutthrough connection
268 for all multi-message initiating connections.
270 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
271 routers. Previously, a multi-recipient message would fail to match the
272 onward-connection opened for the first recipient, and cause its closure.
274 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
275 a timeout on read on a GnuTLS initiating connection, resulting in the
276 initiating connection being dropped. This mattered most when the callout
277 was marked defer_ok. Fix to keep the two timeout-detection methods
280 JH/15 Relax results from ACL control request to enable cutthrough, in
281 unsupported situations, from error to silently (except under debug)
282 ignoring. This covers use with PRDR, frozen messages, queue-only and
285 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
287 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
288 metadata, resulting in a crash in free().
290 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
291 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
292 Broken also in d185889f4, with init system revamp.
294 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
295 Previously we abruptly closed the connection after reading a malware-
296 found indication; now we go on to read the "scan ok" response line,
299 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
300 ACL. Previously, a crash would result.
302 JH/19 Speed up macro lookups during configuration file read, by skipping non-
303 macro text after a replacement (previously it was only once per line) and
304 by skipping builtin macros when searching for an uppercase lead character.
306 JH/20 DANE support moved from Experimental to mainline. The Makefile control
307 for the build is renamed.
309 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
310 was allocated for every new TLS startup, meaning one per message. Fix
311 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
313 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
314 reported the original. Fix to report (as far as possible) the ACL
315 result replacing the original.
317 JH/23 Fix memory leak during multi-message connections using STARTTLS under
318 OpenSSL. Certificate information is loaded for every new TLS startup,
319 and the resources needed to be freed.
321 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
323 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
326 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
327 DATA response info to the (existing) per-recipient response info for
328 the "C=" log element. It can have useful tracking info from the
329 destination system. Patch from Simon Arlott.
331 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
332 length value. Previously this would segfault.
334 HS/02 Support Avast multiline protoocol, this allows passing flags to
335 newer versions of the scanner.
337 JH/28 Ensure that variables possibly set during message acceptance are marked
338 dead before release of memory in the daemon loop. This stops complaints
339 about them when the debug_store option is enabled. Discovered specifically
340 for sender_rate_period, but applies to a whole set of variables.
341 Do the same for the queue-runner and queue-list loops, for variables set
342 from spool message files. Do the same for the SMTP per-message loop, for
343 certain variables indirectly set in ACL operations.
345 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
346 as a multi-recipient message from a mailinglist manager). The coding had
347 an arbitrary cutoff number of characters while checking for more input;
348 enforced by writing a NUL into the buffer. This corrupted long / fast
349 input. The problem was exposed more widely when more pipelineing of SMTP
350 responses was introduced, and one Exim system was feeding another.
351 The symptom is log complaints of SMTP syntax error (NUL chars) on the
352 receiving system, and refused recipients seen by the sending system
353 (propating to people being dropped from mailing lists).
354 Discovered and pinpointed by David Carter.
356 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
357 replaced by the ${authresults } expansion.
359 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
361 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
362 allows proper process termination in container environments.
364 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
365 Previously the "final dot" had a newline after it; ensure it is CR,LF.
367 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
368 and "err_perm", deprecated since 4.83 when the RFC-defined words
369 "temperror" and "permerror" were introduced.
371 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
372 transport-filters or DKIM-signing. The restriction was lost in the
373 consolidation of verify-callout and delivery SMTP handling.
374 Extend the restriction to also cover ARC-signing.
376 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
377 in defer=pass mode supply a 450 to the initiator. Previously the message
380 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
381 tls_require_ciphers is used as before.
383 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
384 "pass_unscanned". Only tmpfails from the scanner are written to
385 the paniclog, as they may require admin intervention (permission
386 denied, license issues). Other scanner errors (like decompression
387 bombs) do not cause a paniclog entry.
389 JH/36 Fix reinitialisation of DKIM logging variable between messages.
390 Previously it was possible to log spurious information in receive log
393 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
394 triggered odd behaviour from Outlook Express clients.
396 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
399 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
400 since the IETF WG has not yet settled on that versus the original
401 "bare" representation.
403 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
404 Previously the millisecond value corrupted the output.
405 Fix also for syslog_pid=no and log_selector +pid, for which the pid
406 corrupted the output.
412 JH/01 Rework error string handling in TLS interface so that the caller in
413 more cases is responsible for logging. This permits library-sourced
414 string to be attached to addresses during delivery, and collapses
415 pairs of long lines into single ones.
417 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
418 during configuration. Wildcards are allowed and expanded.
420 JH/02 Rework error string handling in DKIM to pass more info back to callers.
421 This permits better logging.
423 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
424 do not close it down and have the child transport start it up again on
425 the passed-on TCP connection. Instead, proxy the child (and any
426 subsequent ones) for TLS via a unix-domain socket channel. Logging is
427 affected: the continued delivery log lines do not have any DNSSEC, TLS
428 Certificate or OCSP information. TLS cipher information is still logged.
430 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
431 identical IP addresses on different listening ports. Will also affect
434 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
435 add noisy ifdef guards to special-case this sillyness.
436 Patch from Bernd Kuhls.
438 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
439 than 255 are no longer allowed.
441 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
442 Disable the session-cache too, which might reduce our load. Since we
443 currrectly use a new context for every connection, both as server and
444 client, there is no benefit for these.
445 GnuTLS appears to not support tickets server-side by default (we don't
446 call gnutls_session_ticket_enable_server()) but client side is enabled
447 by default on recent versions (3.1.3 +) unless the PFS priority string
450 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
451 <https://reproducible-builds.org/specs/source-date-epoch/>.
453 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
454 the check for any unsuccessful recipients did not notice the limit, and
455 erroneously found still-pending ones.
457 JH/08 Pipeline CHUNKING command and data together, on kernels that support
458 MSG_MORE. Only in-clear (not on TLS connections).
460 JH/09 Avoid using a temporary file during transport using dkim. Unless a
461 transport-filter is involved we can buffer the headers in memory for
462 creating the signature, and read the spool data file once for the
463 signature and again for transmission.
465 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
466 4.77 as the kernel support then wasn't solid, having issues in 64bit
467 mode. Now, it's been long enough. Add support for FreeBSD also.
469 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
470 case where the routing stage had gathered several addresses to send to
471 a host before calling the transport for the first, we previously failed
472 to close down TLS in the old transport process before passing the TCP
473 connection to the new process. The new one sent a STARTTLS command
474 which naturally failed, giving a failed delivery and bloating the retry
475 database. Investigation and fix prototype from Wolfgang Breyha.
477 JH/12 Fix check on SMTP command input synchronisation. Previously there were
478 false-negatives in the check that the sender had not preempted a response
479 or prompt from Exim (running as a server), due to that code's lack of
480 awareness of the SMTP input buffering.
482 PP/04 Add commandline_checks_require_admin option.
483 Exim drops privileges sanely, various checks such as -be aren't a
484 security problem, as long as you trust local users with access to their
485 own account. When invoked by services which pass untrusted data to
486 Exim, this might be an issue. Set this option in main configuration
487 AND make fixes to the calling application, such as using `--` to stop
490 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
491 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
492 responses to those, into a single TLS record each way (this usually means
493 a single packet). As a side issue, smtp_enforce_sync now works on TLS
496 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
497 affects you only if you're dancing at the edge of the param size limits.
498 If you are, and this message makes sense to you, then: raise the
499 configured limit or use OpenSSL 1.1. Nothing we can do for older
502 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
503 cmdline element to get the documented default one. Previously it was
506 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
507 are unsuable for non-message-specific reasons - eg. connection timeout,
508 banner-time rejection.
510 JH/16 Fix logging of delivery remote port, when specified by router, under
513 PP/06 Repair manualroute's ability to take options in any order, even if one
514 is the name of a transport.
517 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
519 JH/17 Change the list-building routines interface to use the expanding-string
520 triplet model, for better allocation and copying behaviour.
522 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
523 Previously it was constructed the first time a possibly-matching string
524 was met in the configuration file input during startup; now it is done
527 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
528 compatible one, to avoid the (poorly documented) possibility of a config
529 file in the working directory redirecting the DB files, possibly correpting
530 some existing file. CVE-2017-10140 assigned for BDB.
532 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
533 cache-hot. Previously, although the result was properly cached, the
534 initial verify call returned a defer.
536 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
537 the main verify for receipient in uncached-mode.
539 JH/22 Retire historical build files to an "unsupported" subdir. These are
540 defined as "ones for which we have no current evidence of testing".
542 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
543 if present. Previously it was ignored.
545 JH/24 Start using specified-initialisers in C structure init coding. This is
546 a C99 feature (it's 2017, so now considered safe).
548 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
549 if was a fixed-sized field and bitmask ops via macros; it is now more
552 PP/07 GitHub PR 56: Apply MariaDB build fix.
553 Patch provided by Jaroslav Škarvada.
555 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
556 during Coverity cleanups [4.87 JH/47]
557 Diagnosis and fix provided by Michael Fischer v. Mollard.
559 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
560 the right size to place the terminating semicolon on its own folded
561 line, the header hash was calculated to an incorrect value thanks to
562 the (relaxed) space the fold became.
564 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
565 and confused the parent.
567 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
568 which could crash as a result. This could lead to undeliverable messages.
570 JH/28 Logging: "next input sent too soon" now shows where input was truncated
573 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
574 matters on fast-turnover and PID-randomising systems, which were getting
575 out-of-order delivery.
577 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
578 a possibly-overlapping copy. The symptom was that "Remote host closed
579 connection in response to HELO" was logged instead of the actual 4xx
582 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
583 Previously only that bufferd was discarded, resulting in SYMTP command
586 JH/32 DKIM: when a message has multiple signatures matching an identity given
587 in dkim_verify_signers, run the dkim acl once for each. Previously only
588 one run was done. Bug 2189.
590 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
591 "panic the current process" to "deliberately defer". The panic log is
592 still written with the problem list name; the mail and reject logs now
593 get a temp-reject line for the message that was being handled, saying
594 something like "domains check lookup or other defer". The SMTP 451
595 message is still "Temporary local problem".
597 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
598 A crafted sequence of BDAT commands could result in in-use memory beeing
599 freed. CVE-2017-16943.
601 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
602 from SMTP input. Previously it was always done; now only done for DATA
603 and not BDAT commands. CVE-2017-16944.
605 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
606 to the message (such as an overlong header line). Previously this was
607 not done and we did not exit BDAT mode. Followon from the previous item
608 though a different problem.
614 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
615 than -2003 did; needs libidn2 in addition to libidn.
617 JH/02 The path option on a pipe transport is now expanded before use.
619 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
620 Patch provided by "Björn", documentation fix added too.
622 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
623 missing a wire-to-host endian conversion.
625 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
626 close after a BDAT command line could be taken as a following command,
627 giving a synch failure. Fix by only checking for synch immediately
628 before acknowledging the chunk.
630 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
631 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
632 macro. Patches provided by Josh Soref.
634 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
635 Previously we did not; the RFC seems ambiguous and VRFY is not listed
636 by IANA as a service extension. However, John Klensin suggests that we
639 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
640 the dkim code may be unix-mode line endings rather than smtp wire-format
641 CRLF, so prepend a CR to any bare LF.
643 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
644 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
646 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
647 the first were themselves being wrongly included in the feed into dkim
648 processing; with most chunk sizes in use this resulted in an incorrect
649 body hash calculated value.
651 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
652 DKIM signature block, for verification. Although advised against by
653 standards it is specifically not ruled illegal.
655 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
657 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
658 missing a body hash (the bh= tag).
660 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
661 It seems that HAProxy sends the Proxy Protocol information in clear and
662 only then does a TLS startup, so do the same.
664 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
665 TCP connections (such as for Spamd) unless the daemon successfully set
666 Fast Open mode on its listening sockets. This fixes breakage seen on
667 too-old kernels or those not configured for Fast Open, at the cost of
668 requiring both directions being enabled for TFO, and TFO never being used
669 by non-daemon-related Exim processes.
671 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
672 endings, at least on the first header line. Try to canonify any that get
673 past that check, despite the cost.
675 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
676 now limited to an arbitrary five deep, while parsing addresses with the
677 strip_excess_angle_brackets option enabled.
679 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
680 instead leave the unprompted TLS handshake in socket buffer for the
681 TLS library to consume.
683 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
685 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
687 JH/16 Drop variables when they go out of scope. Memory management drops a whole
688 region in one operation, for speed, and this leaves assigned pointers
689 dangling. Add checks run only under the testsuite which checks all
690 variables at a store-reset and panics on a dangling pointer; add code
691 explicitly nulling out all the variables discovered. Fixes one known
692 bug: a transport crash, where a dangling pointer for $sending_ip_address
693 originally assigned in a verify callout, is re-used.
695 PP/06 Drop '.' from @INC in various Perl scripts.
697 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
699 PP/08 Reduce a number of compilation warnings under clang; building with
700 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
701 should be warning-free.
703 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
705 HS/01 Fix portability problems introduced by PP/08 for platforms where
706 realloc(NULL) is not equivalent to malloc() [SunOS et al].
708 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
709 chunk. This allows us to accept broken chunked messages. We need a more
710 general solution here.
712 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
713 already-broken messages in the queue.
715 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
717 JH/19 Fix reference counting bug in routing-generated-address tracking.
723 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
724 supports it and a size is available (ie. the sending peer gave us one).
726 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
727 years of being deprecated). The replacements are the ACLs
728 acl_smtp_mime and acl_not_smtp_mime.
730 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
731 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
732 or even an in-clear connection were permitted. Now, if the host lookup
733 was dnssec and dane was requested then the host is only used if the
734 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
735 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
736 if one fails this test.
737 This means that a poorly-configured remote DNS will make it incommunicado;
738 but it protects against a DNS-interception attack on it.
740 JH/04 Bug 1810: make continued-use of an open smtp transport connection
741 non-noisy when a race steals the message being considered.
743 JH/05 If main configuration option tls_certificate is unset, generate a
744 self-signed certificate for inbound TLS connections.
746 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
747 in rewrites and routers.
749 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
750 and logged a warning sing 4.83; now they are a configuration file error.
752 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
753 (lacking @domain). Apply the same qualification processing as RCPT.
755 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
757 JH/10 Support ${sha256:} applied to a string (as well as the previous
760 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
761 a cutthrough deliver is pending, as we always want to make a connection.
762 This also avoids re-routing the message when later placing the cutthrough
763 connection after a verify cache hit.
764 Do not update it with the verify result either.
766 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
767 when routing results in more than one destination address.
769 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
770 signing (which inhibits the cutthrough capability). Previously only
771 the presence of an option was tested; now an expansion evaluating as
772 empty is permissible (obviously it should depend only on data available
773 when the cutthrough connection is made).
775 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
776 the relevant preceding SMTP command did not note the pipelining mode.
778 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
779 Previously they were not counted.
781 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
782 as one having no matching records. Previously we deferred the message
783 that needed the lookup.
785 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
786 distinguished as "(=".
788 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
789 for missing MX records. Previously it only worked for missing A records.
791 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
793 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
794 after the data-go-ahead and data-ack. Patch from Jason Betts.
796 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
797 even for a "none" policy. Patch from Tony Meyer.
799 JH/22 Fix continued use of a connection for further deliveries. If a port was
800 specified by a router, it must also match for the delivery to be
803 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
804 When one of the recipients of a message was unsuitable for the connection
805 (has no matching addresses), we lost track of needing to mark it
806 deferred. As a result mail would be lost.
808 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
810 JH/25 Decoding ACL controls is now done using a binary search; the source code
811 takes up less space and should be simpler to maintain. Merge the ACL
812 condition decode tables also, with similar effect.
814 JH/26 Fix problem with one_time used on a redirect router which returned the
815 parent address unchanged. A retry would see the parent address marked as
816 delivered, so not attempt the (identical) child. As a result mail would
819 JH/27 Fix a possible security hole, wherein a process operating with the Exim
820 UID can gain a root shell. Credit to http://www.halfdog.net/ for
821 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
824 JH/28 Enable {spool,log} filesystem space and inode checks as default.
825 Main config options check_{log,spool}_{inodes,space} are now
826 100 inodes, 10MB unless set otherwise in the configuration.
828 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
829 Previously it only applied to the main-section connection policy
832 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
834 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
835 by me. Added RFC7919 DH primes as an alternative.
837 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
838 are not in the system include path.
840 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
841 GnuTLS, when a session startup failed (eg because the client disconnected)
842 Exim did stdio operations after fclose. This was exposed by a recent
843 change which nulled out the file handle after the fclose.
845 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
846 signed directly by the cert-signing cert, rather than an intermediate
847 OCSP-signing cert. This is the model used by LetsEncrypt.
849 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
851 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
852 an incoming connection.
854 HS/02 Bug 1802: Do not half-close the connection after sending a request
857 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
858 fallback to "prime256v1".
860 JH/34 SECURITY: Use proper copy of DATA command in error message.
861 Could leak key material. Remotely exploitable. CVE-2016-9963.
867 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
868 and 3.4.4 - once the server is enabled to respond to an OCSP request
869 it does even when not requested, resulting in a stapling non-aware
870 client dropping the TLS connection.
872 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
873 support variable-length bit vectors. No functional change.
875 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
876 The I= interface field on outgoing lines is now after the H= remote
877 host field, same as incoming lines. There is a separate
878 outgoing_interface log selector which allows you to disable the
881 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
882 If not running log_selector +smtp_connection the mainlog would be held
883 open indefinitely after a "too many connections" event, including to a
884 deleted file after a log rotate. Leave the per net connection logging
885 leaving it open for efficiency as that will be quickly detected by the
886 check on the next write.
888 HS/01 Bug 1671: Fix post transport crash.
889 Processing the wait-<transport> messages could crash the delivery
890 process if the message IDs didn't exist for some reason. When
891 using 'split_spool_directory=yes' the construction of the spool
892 file name failed already, exposing the same netto behaviour.
894 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
895 mime_regex ACL conditions.
897 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
898 to DSN fail messages (bounces): remote IP, remote greeting, remote response
899 to HELO, local diagnostic string.
901 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
902 log line to debug. Even when configured with a tls authenticator many
903 client connections are expected to not authenticate in this way, so
904 an authenticate fail is not an error.
906 HS/02 Add the Exim version string to the process info. This way exiwhat
907 gives some more detail about the running daemon.
909 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
910 matter for fast-change records such as DNSBLs.
912 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
913 retry record, even if constant. There may be multiple transports with
914 different interface settings and the retry behaviour needs to be kept
917 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
919 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
921 JH/10 Bug 840: fix log_defer_output option of pipe transport
923 JH/11 Bug 830: use same host for all RCPTS of a message, even under
924 hosts_randomize. This matters a lot when combined with mua_wrapper.
926 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
927 ${quote_pgsql:<string>} operator.
929 JH/13 Bug 1708: avoid misaligned access in cached lookup.
931 JH/14 Change header file name for freeradius-client. Relevant if compiling
932 with Radius support; from the Gentoo tree and checked under Fedora.
934 JH/15 Bug 1712: Introduce $prdr_requested flag variable
936 JH/16 Bug 1714: Permit an empty string as expansion result for transport
937 option transport_filter, meaning no filtering.
939 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
941 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
942 defaults to "*" (all hosts). The variable is now available when not built
943 with TLS, default unset, mainly to enable keeping the testsuite sane.
944 If a server certificate is not supplied (via tls_certificate) an error is
945 logged, and clients will find TLS connections fail on startup. Presumably
946 they will retry in-clear.
947 Packagers of Exim are strongly encouraged to create a server certificate
948 at installation time.
950 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
951 with the $config_file variable.
953 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
954 in transport context, after the attempt, and per-recipient. The latter type
955 is per host attempted. The event data is the error message, and the errno
956 information encodes the lookup type (A vs. MX) used for the (first) host,
957 and the trailing two digits of the smtp 4xx response.
959 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
960 to write to mainlog (or rejectlog, paniclog) in the window between file
961 creation and permissions/ownership being changed. Particularly affects
962 installations where exicyclog is run as root, rather than exim user;
963 result is that the running daemon panics and dies.
965 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
967 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
968 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
969 "pri" and "weight". Note that the previous implicit priority given by the
970 list order is no longer honoured.
972 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
975 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
976 by defining SUPPORT_SOCKS.
978 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
979 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
980 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
981 variables are renamed to proxy_{local,external}_{address,port}.
983 JH/27 Move Internationalisation support from Experimental to mainline, enabled
984 for a build by defining SUPPORT_I18N
986 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
987 of the query string, and make ${quote_redis:} do that quoting.
989 JH/29 Move Events support from Experimental to mainline, enabled by default
990 and removable for a build by defining DISABLE_EVENT.
992 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
994 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
995 cached by the daemon.
997 JH/32 Move Redis support from Experimental to mainline, enabled for a build
998 by defining LOOKUP_REDIS. The libhiredis library is required.
1000 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1001 keys are given for lookup.
1003 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1004 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1005 only supported when built with TLS support. The PolarSSL SHA routines
1006 are still used when the TLS library is too old for convenient support.
1008 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1009 openssl_options), for security. OpenSSL forces this from version 1.1.0
1010 server-side so match that on older versions.
1012 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1013 allocation for $value could be released as the expansion processing
1014 concluded, but leaving the global pointer active for it.
1016 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1017 and to use the domains and local_parts ACL conditions.
1019 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1020 incorrectly not doubled on cutthrough transmission, hence seen as a
1021 body-termination at the receiving system - resulting in truncated mails.
1022 Commonly the sender saw a TCP-level error, and retransmitted the message
1023 via the normal store-and-forward channel. This could result in duplicates
1024 received - but deduplicating mailstores were liable to retain only the
1025 initial truncated version.
1027 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1029 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1031 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1032 we're in there, support oversigning also; bug 1309.
1034 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1036 HS/04 Add support for keep_environment and add_environment options.
1038 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1039 either intentional arithmetic overflow during PRNG, or testing config-
1042 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1043 delivery resulted in actual delivery. Cancel cutthrough before DATA
1046 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1047 rejects a recipient: pass the reject to the originator.
1049 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1050 Many were false-positives and ignorable, but it's worth fixing the
1053 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1054 for the new environment-manipulation done at startup. Move the routines
1055 from being local to tls.c to being global via the os.c file.
1057 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1058 an extract embedded as result-arg for a map, the first arg for extract
1059 is unavailable so we cannot tell if this is a numbered or keyed
1060 extraction. Accept either.
1066 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1069 JH/02 The smtp transport option "multi_domain" is now expanded.
1071 JH/03 The smtp transport now requests PRDR by default, if the server offers
1074 JH/04 Certificate name checking on server certificates, when exim is a client,
1075 is now done by default. The transport option tls_verify_cert_hostnames
1076 can be used to disable this per-host. The build option
1077 EXPERIMENTAL_CERTNAMES is withdrawn.
1079 JH/05 The value of the tls_verify_certificates smtp transport and main options
1080 default to the word "system" to access the system default CA bundle.
1081 For GnuTLS, only version 3.0.20 or later.
1083 JH/06 Verification of the server certificate for a TLS connection is now tried
1084 (but not required) by default. The verification status is now logged by
1085 default, for both outbound TLS and client-certificate supplying inbound
1088 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1091 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1092 Status Notification (bounce) messages are now MIME format per RFC 3464.
1093 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1094 under the control of the dsn_advertise_hosts option, and routers may
1095 have a dsn_lasthop option.
1097 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1098 default, modifiable by a malware= option. The list separator for
1099 the options can now be changed in the usual way. Bug 68.
1101 JH/10 The smtp_receive_timeout main option is now expanded before use.
1103 JH/11 The incoming_interface log option now also enables logging of the
1104 local interface on delivery outgoing connections.
1106 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1107 if the interface and destination host and port all match.
1109 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1112 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1113 Patch from Andrew Lewis.
1115 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1116 now supports optional time-restrictions, weighting, and priority
1117 modifiers per server. Patch originally by <rommer@active.by>.
1119 JH/16 The spamd_address main option now supports a mixed list of local
1120 and remote servers. Remote servers can be IPv6 addresses, and
1121 specify a port-range.
1123 JH/17 Bug 68: The spamd_address main option now supports an optional
1124 timeout value per server.
1126 JH/18 Bug 1581: Router and transport options headers_add/remove can
1127 now have the list separator specified.
1129 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1132 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1135 JH/21 Support for the A6 type of dns record is withdrawn.
1137 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1138 rather than the verbs used.
1140 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1141 from 255 to 1024 chars.
1143 JH/24 Verification callouts now attempt to use TLS by default.
1145 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1146 are generic router options now. The defaults didn't change.
1148 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1149 Original patch from Alexander Shikoff, worked over by JH.
1151 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1152 files if ZCAT_COMMAND is not executable.
1154 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1156 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1158 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1159 Normally benign, it bites when the pair was led to by a CNAME;
1160 modern usage is to not canonicalize the domain to a CNAME target
1161 (and we were inconsistent anyway for A-only vs AAAA+A).
1163 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1165 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1166 when evaluating $sender_host_dnssec.
1168 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1169 $sender_helo_dnssec variable.
1171 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1173 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1175 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1177 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1178 documented as working, but never had. Support all but $spam_report.
1180 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1181 added for tls authenticator.
1183 HS/03 Add perl_taintmode main config option
1189 TL/01 When running the test suite, the README says that variables such as
1190 no_msglog_check are global and can be placed anywhere in a specific
1191 test's script, however it was observed that placement needed to be near
1192 the beginning for it to behave that way. Changed the runtest perl
1193 script to read through the entire script once to detect and set these
1194 variables, reset to the beginning of the script, and then run through
1195 the script parsing/test process like normal.
1197 TL/02 The BSD's have an arc4random API. One of the functions to induce
1198 adding randomness was arc4random_stir(), but it has been removed in
1199 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1200 function when detected.
1202 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1203 cause callback expansion.
1205 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1206 syntax errors in an expansion can be treated as a string instead of
1207 logging or causing an error, due to the internal use of bool_lax
1208 instead of bool when processing it.
1210 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1211 server certificates when making smtp deliveries.
1213 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1215 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1217 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1219 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1220 Merged patch from Sebastian Wiedenroth.
1222 JH/05 Fix results-pipe from transport process. Several recipients, combined
1223 with certificate use, exposed issues where response data items split
1224 over buffer boundaries were not parsed properly. This eventually
1225 resulted in duplicates being sent. This issue only became common enough
1226 to notice due to the introduction of connection certificate information,
1227 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1229 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1230 size buffer was used, resulting in syntax errors when an expansion
1233 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1234 version 3.3.6 or later.
1236 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1237 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1238 both become "event_action", the variables become $event_name, $event_data
1239 and $event_defer_errno. There is a new variable $verify_mode, usable in
1240 routers, transports and related events. The tls:cert event is now also
1241 raised for inbound connections, if the main configuration event_action
1244 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1245 early OCSP support, but no stapling (appears to be less than 1.0.0).
1247 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1248 server certificate names available under the smtp transport option
1249 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1252 JH/10 Time-related extraction expansions from certificates now use the main
1253 option "timezone" setting for output formatting, and are consistent
1254 between OpenSSL and GnuTLS compilations. Bug 1541.
1256 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1257 encoded parameter in the incoming message. Bug 1558.
1259 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1260 include certificate info, eximon was claiming there were spoolfile
1263 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1265 JH/14 Log delivery-related information more consistently, using the sequence
1266 "H=<name> [<ip>]" wherever possible.
1268 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1269 are problematic for Debian distribution, omit them from the release
1272 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1274 JH/16 Fix string representation of time values on 64bit time_t architectures.
1277 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1278 output list separator was used.
1283 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1284 checkers that were complaining about end of non-void function with no
1287 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1288 This was a regression introduced in 4.83 by another bugfix.
1290 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1292 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1293 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1299 TF/01 Correctly close the server side of TLS when forking for delivery.
1301 When a message was received over SMTP with TLS, Exim failed to clear up
1302 the incoming connection properly after forking off the child process to
1303 deliver the message. In some situations the subsequent outgoing
1304 delivery connection happened to have the same fd number as the incoming
1305 connection previously had. Exim would try to use TLS and fail, logging
1306 a "Bad file descriptor" error.
1308 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1309 utilities have not been installed.
1311 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1312 temporary space as the ACL may create new global variables.
1314 TL/01 LDAP support uses per connection or global context settings, depending
1315 upon the detected version of the libraries at build time.
1317 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1318 to extract and use the src ip:port in logging and expansions as if it
1319 were a direct connection from the outside internet. PPv2 support was
1320 updated based on HAProxy spec change in May 2014.
1322 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1324 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1325 Properly escape header and check for NULL return.
1327 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1330 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1332 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1333 characters in header names, implemented as a verify condition.
1334 Contributed by Michael Fischer v. Mollard.
1336 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1337 results permerror and temperror. Previous values are deprecated but
1338 still accepted. In a future release, err_perm and err_temp will be
1339 completely removed, which will be a backward incompatibility if the
1340 ACL tests for either of these two old results. Patch contributed by
1341 user bes-internal on the mailing list.
1343 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1345 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1346 selectors, in both main and reject logs.
1348 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1351 JH/07 Add malware type "sock" for talking to simple daemon.
1353 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1355 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1356 routers/transports under cutthrough routing.
1358 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1359 numbers. Touch up "bool" conditional to keep the same definition.
1361 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1363 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1365 JH/12 Expand items in router/transport headers_add or headers_remove lists
1366 individually rather than the list as a whole. Bug 1452.
1368 Required for reasonable handling of multiple headers_ options when
1369 they may be empty; requires that headers_remove items with embedded
1370 colons must have them doubled (or the list-separator changed).
1372 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1373 view the policy declared in the DMARC record. Currently, $dmarc_status
1374 is a combined value of both the record presence and the result of the
1377 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1379 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1380 dnslookup router and the smtp transport (applying to the forward
1383 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1384 of ldap servers used for a specific lookup. Patch provided by Heiko
1387 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1388 New variable $lookup_dnssec_authenticated for observability.
1390 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1391 Patch submitted by Lars Timman.
1393 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1395 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1396 Requires trusted mode and valid format message id, aborts otherwise.
1397 Patch contributed by Heiko Schlichting.
1399 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1400 certextract with support for various fields. Bug 1358.
1402 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1403 is requested by default, modifiable by smtp transport option
1406 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1407 operate on certificate variables to give certificate fingerprints
1408 Also new ${sha256:cert_variable}.
1410 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1412 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1413 Christian Aistleitner.
1415 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1417 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1418 file. Patch from Wolfgang Breyha.
1420 JH/25 Expand the coverage of the delivery $host and $host_address to
1421 client authenticators run in verify callout. Bug 1476.
1423 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1424 align with daemon_smtp_ports. Bug 72.
1426 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1427 support and error reporting did not work properly.
1429 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1430 and is readable. Patch from Andrew Colin Kissa.
1432 TL/14 Enhance documentation of ${run expansion and how it parses the
1433 commandline after expansion, particularly in the case when an
1434 unquoted variable expansion results in an empty value.
1436 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1438 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1439 by renaming away from it. Addresses GH issue 15, reported by
1442 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1444 TL/15 SECURITY: prevent double expansion in math comparison functions
1445 (can expand unsanitized data). Not remotely exploitable.
1452 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1454 PP/02 Make -n do something, by making it not do something.
1455 When combined with -bP, the name of an option is not output.
1457 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1460 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1461 $sender_host_name and config options to manage this, and basic check
1464 PP/05 DSCP support for outbound connections and control modifier for inbound.
1466 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1467 (Only plugin which currently uses this is kerberos4, which nobody should
1468 be using, but we should make it available and other future plugins might
1469 conceivably use it, even though it would break NAT; stuff *should* be
1470 using channel bindings instead).
1472 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1473 name; added for Sendmail compatibility; requires admin caller.
1474 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1475 just ignore it); requires trusted caller.
1476 Also parse but ignore: -Ac -Am -X<logfile>
1479 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1481 TL/02 Add +smtp_confirmation as a default logging option.
1483 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1484 Patch by Magnus Holmgren from 2007-02-20.
1486 TL/04 Bugzilla 1281 - Spec typo.
1487 Bugzilla 1283 - Spec typo.
1488 Bugzilla 1290 - Spec grammar fixes.
1490 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1492 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1494 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1495 mailing list by by Dmitry Isaikin.
1497 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1499 JH/02 Support "G" suffix to numbers in ${if comparisons.
1501 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1503 NM/01 Bugzilla 1197 - Spec typo
1504 Bugzilla 1196 - Spec examples corrections
1506 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1508 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1509 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1512 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1513 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1515 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1516 "acl {{name}{arg}...}", and optional args on acl condition
1519 JH/05 Permit multiple router/transport headers_add/remove lines.
1521 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1523 JH/07 Avoid using a waiting database for a single-message-only transport.
1524 Performance patch from Paul Fisher. Bugzilla 1262.
1526 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1529 JH/09 Add $headers_added variable, with content from use of ACL modifier
1530 add_header (but not yet added to the message). Bugzilla 199.
1532 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1533 Pulled from Bugzilla 817 by Wolfgang Breyha.
1535 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1537 (nb: this is the same fix as in Exim 4.80.1)
1539 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1542 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1545 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1547 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1548 advertise SMTP AUTH mechanism to us, instead of a generic
1549 protocol violation error. Also, make Exim more robust to bad
1550 data from the Dovecot auth socket.
1552 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1554 When a queue runner is handling a message, Exim first routes the
1555 recipient addresses, during which it prunes them based on the retry
1556 hints database. After that it attempts to deliver the message to
1557 any remaining recipients. It then updates the hints database using
1560 So if a recipient address works intermittently, it can get repeatedly
1561 deferred at routing time. The retry hints record remains fresh so the
1562 address never reaches the final cutoff time.
1564 This is a fairly common occurrence when a user is bumping up against
1565 their storage quota. Exim had some logic in its local delivery code
1566 to deal with this. However it did not apply to per-recipient defers
1567 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1569 This change adds a proper retry rule check during routing so that the
1570 final cutoff time is checked against the message's age. We only do
1571 this check if there is an address retry record and there is not a
1572 domain retry record; this implies that previous attempts to handle
1573 the address had the retry_use_local_parts option turned on. We use
1574 this as an approximation for the destination being like a local
1575 delivery, as in LMTP.
1577 I suspect this new check makes the old local delivery cutoff check
1578 redundant, but I have not verified this so I left the code in place.
1580 TF/02 Correct gecos expansion when From: is a prefix of the username.
1582 Test 0254 submits a message to Exim with the header
1586 When I ran the test suite under the user fanf2, Exim expanded
1587 the header to contain my full name, whereas it should have added
1588 a Resent-Sender: header. It erroneously treats any prefix of the
1589 username as equal to the username.
1591 This change corrects that bug.
1593 GF/01 DCC debug and logging tidyup
1594 Error conditions log to paniclog rather than rejectlog.
1595 Debug lines prefixed by "DCC: " to remove any ambiguity.
1597 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1599 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1600 Bug spotted by Jeremy Harris; was flawed since initial commit.
1601 Would have resulted in OCSP responses post-SNI triggering an Exim
1602 NULL dereference and crash.
1604 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1606 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1607 Bug detection, analysis and fix by Samuel Thibault.
1608 Bugzilla 1331, Debian bug #698092.
1610 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1612 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1613 Server implementation by Todd Lyons, client by JH.
1614 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1615 config variable "prdr_enable" controls whether the server
1616 advertises the facility. If the client requests PRDR a new
1617 acl_data_smtp_prdr ACL is called once for each recipient, after
1618 the body content is received and before the acl_smtp_data ACL.
1619 The client is controlled by both of: a hosts_try_prdr option
1620 on the smtp transport, and the server advertisement.
1621 Default client logging of deliveries and rejections involving
1622 PRDR are flagged with the string "PRDR".
1624 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1625 fclose(). Diagnosis by Todd Lyons.
1627 PP/17 Update configure.default to handle IPv6 localhost better.
1628 Patch by Alain Williams (plus minor tweaks).
1631 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1632 This is now consistent with GnuTLS, and is now documented: the
1633 previous undocumented portable approach to treating the option as
1634 unset was to force an expansion failure. That still works, and
1635 an empty string is now equivalent.
1637 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1638 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1639 not performing validation itself.
1641 PP/20 Added force_command boolean option to pipe transport.
1642 Patch from Nick Koston, of cPanel Inc.
1644 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1647 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1649 PP/21 Fix eximon continuous updating with timestamped log-files.
1650 Broken in a format-string cleanup in 4.80, missed when I repaired the
1651 other false fix of the same issue.
1652 Report and fix from Heiko Schlichting.
1655 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1656 Report from Prashanth Katuri.
1658 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1659 It's SecureTransport, so affects any MacOS clients which use the
1660 system-integrated TLS libraries, including email clients.
1662 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1663 using a MIME ACL for non-SMTP local injection.
1664 Report and assistance in diagnosis by Warren Baker.
1666 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1668 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1670 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1671 last id that failed so it may be referenced in subsequent ACL's.
1673 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1676 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1677 ldap library initialization, allowing self-signed CA's to be
1678 used. Also properly sets require_cert option later in code by
1679 using NULL (global ldap config) instead of ldap handle (per
1680 session). Bug diagnosis and testing by alxgomz.
1682 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1683 the src/util/ subdirectory.
1685 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1686 renamed to Transport Post Delivery Action by Jeremy Harris, as
1689 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1690 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1691 redis_servers = needs to be configured which will be used by the redis
1692 lookup. Patch from Warren Baker, of The Packet Hub.
1694 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1696 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1697 hostname or reverse DNS when processing a host list. Used suggestions
1698 from multiple comments on this bug.
1700 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1702 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1703 Missed a few lines, added it to make the runtest require no keyboard
1706 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1707 contains upper case chars. Make router use caseful_local_part.
1709 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1710 support when GnuTLS has been built with p11-kit.
1716 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1718 This, or similar/improved, will also be change PP/11 of 4.82.
1724 PP/01 Handle short writes when writing local log-files.
1725 In practice, only affects FreeBSD (8 onwards).
1726 Bugzilla 1053, with thanks to Dmitry Isaikin.
1728 NM/01 Bugzilla 949 - Documentation tweak
1730 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1733 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1735 PP/02 Implemented gsasl authenticator.
1737 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1739 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1740 `pkg-config foo` for cflags/libs.
1742 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1743 with rest of GSASL and with heimdal_gssapi.
1745 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1746 `pkg-config foo` for cflags/libs for the TLS implementation.
1748 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1749 properties get this fed in as external SSF. A number of robustness
1750 and debugging improvements to the cyrus_sasl authenticator.
1752 PP/08 cyrus_sasl server now expands the server_realm option.
1754 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1755 Patch by Jeremy Harris.
1757 PP/10 Added dbmjz lookup type.
1759 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1761 PP/12 MAIL args handles TAB as well as SP, for better interop with
1762 non-compliant senders.
1763 Analysis and variant patch by Todd Lyons.
1765 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1766 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1767 Patch from Dirk Mueller <dmueller@suse.com>
1769 PP/13 tls_peerdn now print-escaped for spool files.
1770 Observed some $tls_peerdn in wild which contained \n, which resulted
1771 in spool file corruption.
1773 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1774 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1775 or write after TLS renegotiation, which otherwise led to messages
1778 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1779 as a tracking header (ie: a signed header comes before the signature).
1780 Patch from Wolfgang Breyha.
1782 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1783 comma-sep list; embedded commas doubled.
1785 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1787 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1789 Report and patch from Dmitry Banschikov.
1791 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1792 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1793 failures. We appear to now support TLS1.1+ with Exim.
1795 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1796 lets Exim select keys and certificates based upon TLS SNI from client.
1797 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1798 before an outbound SMTP session. New log_selector, +tls_sni.
1800 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1801 NULL dereference. Report and patch from Alun Jones.
1803 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1804 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1805 Not seeing resolver debug output on NetBSD, but suspect this is a
1806 resolver implementation change.
1808 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1809 Left warnings. Added "eximon gdb" invocation mode.
1811 PP/21 Defaulting "accept_8bitmime" to true, not false.
1813 PP/22 Added -bw for inetd wait mode support.
1815 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1816 locate the relevant includes and libraries. Made this the default.
1818 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1819 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1821 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1822 This may cause build issues on older platforms.
1824 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1825 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1826 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1827 Added SNI support via GnuTLS too.
1828 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1830 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1832 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1833 Applied second patch from Janne, implementing suggestion to default
1834 multiple-strings-in-record handling to match SPF spec.
1836 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1838 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1839 read-only, out of scope).
1840 Patch from Wolfgang Breyha, report from Stuart Northfield.
1842 PP/29 Fix three issues highlighted by clang analyser static analysis.
1843 Only crash-plausible issue would require the Cambridge-specific
1844 iplookup router and a misconfiguration.
1845 Report from Marcin Mirosław.
1847 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1849 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1850 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1851 As part of this, removing so much warning spew let me fix some minor
1852 real issues in debug logging.
1854 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1855 assignment on my part. Fixed.
1857 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1858 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1859 Janne Snabb (who went above and beyond: thank you).
1861 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1862 string otherwise requires a connection and a bunch more work and it's
1863 relatively easy to get wrong. Should also expose TLS library linkage
1866 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1867 64-bit ${eval} (JH/03).
1869 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1870 GNU libc to support some of the 64-bit stuff, should not lead to
1871 conflicts. Defined before os.h is pulled in, so if a given platform
1872 needs to override this, it can.
1874 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1875 protection layer was required, which is not implemented.
1876 Bugzilla 1254, patch from Wolfgang Breyha.
1878 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1879 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1880 tls_dhparam take prime identifiers. Also unbreak combination of
1881 OpenSSL+DH_params+TLSSNI.
1883 PP/39 Disable SSLv2 by default in OpenSSL support.
1889 PP/01 Solaris build fix for Oracle's LDAP libraries.
1890 Bugzilla 1109, patch from Stephen Usher.
1892 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1894 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1897 TF/02 Fix a couple more cases where we did not log the error message
1898 when unlink() failed. See also change 4.74-TF/03.
1900 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1901 lock up or crash if it happened to be inside a call to libc when it
1902 got a SIGUSR1 from exiwhat.
1904 The SIGUSR1 handler appends the current process status to the process
1905 log which is later printed by exiwhat. It used to use the general
1906 purpose logging code to do this, but several functions it calls are
1907 not safe for signals.
1909 The new output code in the SIGUSR1 handler is specific to the process
1910 log, and simple enough that it's easy to inspect for signal safety.
1911 Removing some special cases also simplifies the general logging code.
1912 Removing the spurious timestamps from the process log simplifies
1915 TF/04 Improved ratelimit ACL condition.
1917 The /noupdate option has been deprecated in favour of /readonly which
1918 has clearer semantics. The /leaky, /strict, and /readonly update modes
1919 are mutually exclusive. The update mode is no longer included in the
1920 database key; it just determines when the database is updated. (This
1921 means that when you upgrade Exim will forget old rate measurements.)
1923 Exim now checks that the per_* options are used with an update mode that
1924 makes sense for the current ACL. For example, when Exim is processing a
1925 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1926 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1927 must specify per_mail/readonly. If you omit the update mode it defaults to
1928 /leaky where that makes sense (as before) or /readonly where required.
1930 The /noupdate option is now undocumented but still supported for
1931 backwards compatibility. It is equivalent to /readonly except that in
1932 ACLs where /readonly is required you may specify /leaky/noupdate or
1933 /strict/noupdate which are treated the same as /readonly.
1935 A useful new feature is the /count= option. This is a generalization
1936 of the per_byte option, so that you can measure the throughput of other
1937 aggregate values. For example, the per_byte option is now equivalent
1938 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1940 The per_rcpt option has been generalized using the /count= mechanism
1941 (though it's more complicated than the per_byte equivalence). When it is
1942 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1943 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1944 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1945 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1946 Note that using per_rcpt with a non-readonly update mode in more than
1947 one ACL will cause the recipients to be double-counted. (The per_mail
1948 and per_byte options don't have this problem.)
1950 The handling of very low rates has changed slightly. If the computed rate
1951 is less than the event's count (usually one) then this event is the first
1952 after a long gap. In this case the rate is set to the same as this event's
1953 count, so that the first message of a spam run is counted properly.
1955 The major new feature is a mechanism for counting the rate of unique
1956 events. The new per_addr option counts the number of different
1957 recipients that someone has sent messages to in the last time period. It
1958 behaves like per_rcpt if all the recipient addresses are different, but
1959 duplicate recipient addresses do not increase the measured rate. Like
1960 the /count= option this is a general mechanism, so the per_addr option
1961 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1962 example, measure the rate that a client uses different sender addresses
1963 with the options per_mail/unique=$sender_address. There are further
1964 details in the main documentation.
1966 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1968 TF/06 Removed a few PCRE remnants.
1970 TF/07 Automatically extract Exim's version number from tags in the git
1971 repository when doing development or release builds.
1973 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1974 Bugzilla 879. Patch from Paul Fisher.
1976 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1977 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1980 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1982 PP/05 Variable $av_failed, true if the AV scanner deferred.
1983 Bugzilla 1078. Patch from John Horne.
1985 PP/06 Stop make process more reliably on build failure.
1986 Bugzilla 1087. Patch from Heiko Schlittermann.
1988 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1989 Bugzilla 1089. Patch from Heiko Schlittermann.
1991 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1992 Bugzilla 1131. Patch from Holger Weiß.
1994 PP/09 Handle IPv6 addresses with SPF.
1995 Bugzilla 860. Patch from Wolfgang Breyha.
1997 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1999 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2002 PP/11 match_* no longer expand right-hand-side by default.
2003 New compile-time build option, EXPAND_LISTMATCH_RHS.
2004 New expansion conditions, "inlist", "inlisti".
2006 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2008 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2010 PP/14 fix log_write() format string regression from TF/03.
2011 Bugzilla 1152. Patch from Dmitry Isaikin.
2017 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2019 PP/02 Harmonised TLS library version reporting; only show if debugging.
2020 Layout now matches that introduced for other libraries in 4.74 PP/03.
2022 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2024 PP/04 New "dns_use_edns0" global option.
2026 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2029 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2030 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2032 TK/01 Updated PolarSSL code to 0.14.2.
2033 Bugzilla 1097. Patch from Andreas Metzler.
2035 PP/07 Catch divide-by-zero in ${eval:...}.
2036 Fixes bugzilla 1102.
2038 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2041 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2042 format-string attack -- SECURITY: remote arbitrary code execution.
2044 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2045 time unintentionally subject to list matching rules, letting the header
2046 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2047 arbitrary string expansion). This allowed for information disclosure.
2049 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2050 INT_MIN/-1 -- value coerced to INT_MAX.
2056 NM/01 Workaround for PCRE version dependency in version reporting
2059 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2060 This fixes portability to compilers other than gcc, notably
2061 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2063 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2064 makefiles for portability to HP-UX and POSIX correctness.
2066 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2067 Also via indented variable definition in the Makefile.
2068 (Debugging by Oliver Heesakkers).
2070 PP/02 Restore caching of spamd results with expanded spamd_address.
2071 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2073 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2074 Improves build reliability. Fix from: Frank Elsner
2076 NM/02 Fix wide character breakage in the rfc2047 coding
2077 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2079 NM/03 Allow underscore in dnslist lookups
2080 Fixes bug 1026. Patch from Graeme Fowler
2082 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2083 Code patches from Adam Ciarcinski of NetBSD.
2085 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2088 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2089 is logged, to avoid truncation. Patch from John Horne.
2091 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2092 Patch from Jakob Hirsch.
2094 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2095 SQL string expansion failure details.
2096 Patch from Andrey Oktyabrski.
2098 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2099 Patch from Simon Arlott.
2101 PP/09 New lookups functionality failed to compile on old gcc which rejects
2102 extern declarations in function scope.
2103 Patch from Oliver Fleischmann
2105 PP/10 Use sig_atomic_t for flags set from signal handlers.
2106 Check getgroups() return and improve debugging.
2107 Fixed developed for diagnosis in bug 927 (which turned out to be
2110 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2111 Patch from Mark Zealey.
2113 PP/12 Bugzilla 1056: Improved spamd server selection.
2114 Patch from Mark Zealey.
2116 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2117 Based on patch from Heiko Schlittermann.
2119 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2120 Patch from Uwe Doering, sign-off by Michael Haardt.
2122 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2123 variable declaration deep within a block. Bug and patch from
2126 PP/15 lookups-Makefile IRIX compatibility coercion.
2128 PP/16 Make DISABLE_DKIM build knob functional.
2130 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2131 Patch by Simon Arlott
2133 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2134 variable argument macros. Our copy now differs from upstream.
2140 TF/01 Failure to get a lock on a hints database can have serious
2141 consequences so log it to the panic log.
2143 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2144 controlled using the smtp_confirmation log selector.
2146 TF/03 Include the error message when we fail to unlink a spool file.
2148 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2149 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2150 for maintaining out-of-tree patches for some time.
2152 PP/01 Bugzilla 139: Documentation and portability issues.
2153 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2154 Handle per-OS dynamic-module compilation flags.
2156 PP/02 Let /dev/null have normal permissions.
2157 The 4.73 fixes were a little too stringent and complained about the
2158 permissions on /dev/null. Exempt it from some checks.
2159 Reported by Andreas M. Kirchwitz.
2161 PP/03 Report version information for many libraries, including
2162 Exim version information for dynamically loaded libraries. Created
2163 version.h, now support a version extension string for distributors
2164 who patch heavily. Dynamic module ABI change.
2166 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2167 privilege escalation vulnerability whereby the Exim run-time user
2168 can cause root to append content of the attacker's choosing to
2171 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2174 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2175 If dropping privileges for untrusted macros, we disabled normal logging
2176 on the basis that it would fail; for the Exim run-time user, this is not
2177 the case, and it resulted in successful deliveries going unlogged.
2178 Fixed. Reported by Andreas Metzler.
2184 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2185 only prepend for the Resent-* case. Fixes regression introduced in
2186 Exim 4.70 by NM/22 for Bugzilla 607.
2188 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2189 increasing numbers of administrators be bitten by this.
2191 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2193 PP/03 Bugzilla 994: added openssl_options main configuration option.
2195 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2197 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2199 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2201 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2202 without a peer certificate, leading to a segfault because of an
2203 assumption that peers always have certificates. Be a little more
2204 paranoid. Problem reported by Martin Tscholak.
2206 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2207 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2208 NB: ClamAV planning to remove STREAM in "middle of 2010".
2209 CL also introduces -bmalware, various -d+acl logging additions and
2210 more caution in buffer sizes.
2212 PP/09 Implemented reverse_ip expansion operator.
2214 PP/10 Bugzilla 937: provide a "debug" ACL control.
2216 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2218 PP/12 Bugzilla 973: Implement --version.
2220 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2222 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2224 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2226 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2227 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2228 ignore trailing whitespace.
2230 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2232 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2235 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2236 Notification from Dr Andrew Aitchison.
2238 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2239 ExtendedDetectionInfo response format.
2240 Notification from John Horne.
2242 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2245 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2246 XSL and documented dependency on system catalogs, with examples of how
2249 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2252 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2253 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2254 configuration file which is writeable by the Exim user or group.
2256 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2257 of configuration files to cover files specified with the -C option if
2258 they are going to be used with root privileges, not just the default
2261 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2262 option (effectively making it always true).
2264 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2265 files to be used while preserving root privileges.
2267 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2268 that rogue child processes cannot use them.
2270 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2271 run-time user, instead of root.
2273 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2274 Exim run-time user without dropping privileges.
2276 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2277 result string, instead of calling string_vformat() twice with the same
2280 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2281 for other users. Others should always drop root privileges if they use
2282 -C on the command line, even for a whitelisted configure file.
2284 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2286 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2292 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2293 $data_path, and $header_path variables; fixed documentation bugs and
2296 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2297 exipick to access non-standard spools, including the "frozen" queue
2300 NM/01 Bugzilla 965: Support mysql stored procedures.
2301 Patch from Alain Williams
2303 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2305 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2306 Patch from Andreas Metzler
2308 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2309 Patch from Kirill Miazine
2311 NM/05 Bugzilla 671: Added umask to procmail example.
2313 JJ/03 installed exipick 20100323.0, fixing doc bug
2315 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2316 directory. Notification and patch from Dan Rosenberg.
2318 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2320 TK/02 Improve log output when DKIM signing operation fails.
2322 MH/01 Treat the transport option dkim_domain as a colon separated
2323 list, not as a single string, and sign the message with each element,
2324 omitting multiple occurences of the same signer.
2326 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2327 Bugzilla 985, 986. Patch by Simon Arlott
2329 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2330 Patch by Simon Arlott
2332 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2333 MBX locking. Notification from Dan Rosenberg.
2339 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2341 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2343 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2345 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2347 NM/04 Bugzilla 915: Flag broken perl installation during build.
2353 TK/01 Added patch by Johannes Berg that expands the main option
2354 "spamd_address" if it starts with a dollar sign.
2356 TK/02 Write list of recipients to X-Envelope-Sender header when building
2357 the mbox-format spool file for content scanning (suggested by Jakob
2360 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2361 (http://www.dcc-servers.net/) support via dccifd. Activated by
2362 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2364 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2365 by Mark Daniel Reidel <mr@df.eu>.
2367 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2368 When building exim an external PCRE library is now needed -
2369 PCRE is a system library on the majority of modern systems.
2370 See entry on PCRE_LIBS in EDITME file.
2372 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2373 conversation. Added nologin parameter to request.
2374 Patch contributed by Kirill Miazine.
2376 TF/01 Do not log submission mode rewrites if they do not change the address.
2378 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2380 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2381 log files in place. Contributed by Roberto Lima.
2383 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2385 TF/03 Bugzilla 615: When checking the local_parts router precondition
2386 after a local_part_suffix or local_part_prefix option, Exim now
2387 does not use the address's named list lookup cache, since this
2388 contains cached lookups for the whole local part.
2390 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2391 Robert Millan. Documentation is in experimental-spec.txt.
2393 TF/04 Bugzilla 668: Fix parallel build (make -j).
2395 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2397 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2398 Patch provided by Jan Srzednicki.
2400 TF/05 Leading white space used to be stripped from $spam_report which
2401 wrecked the formatting. Now it is preserved.
2403 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2404 that they are available at delivery time.
2406 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2408 TF/08 TLS error reporting now respects the incoming_interface and
2409 incoming_port log selectors.
2411 TF/09 Produce a more useful error message if an SMTP transport's hosts
2412 setting expands to an empty string.
2414 NM/06 Bugzilla 744: EXPN did not work under TLS.
2415 Patch provided by Phil Pennock.
2417 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2418 Patch provided by Richard Godbee.
2420 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2421 acl_smtp_notquit, added index entry.
2423 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2424 Patch provided by Eugene Bujak.
2426 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2427 accept(). Patch provided by Maxim Dounin.
2429 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2430 Patch provided by Phil Pennock.
2432 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2434 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2435 Patch provided by Brad "anomie" Jorsch.
2437 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2438 Patch provided by Dean Brooks.
2440 TK/05 Add native DKIM support (does not depend on external libraries).
2442 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2443 Patch provided by Graeme Fowler.
2445 NM/16 Bugzilla 851: Documentation example syntax fix.
2447 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2449 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2452 NM/19 Bugzilla 745: TLS version reporting.
2453 Patch provided by Phil Pennock.
2455 NM/20 Bugzilla 167: bool: condition support.
2456 Patch provided by Phil Pennock.
2458 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2459 clients. Patch provided by Phil Pennock.
2461 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2462 Patch provided by Brad "anomie" Jorsch.
2464 NM/23 Bugzilla 687: Fix misparses in eximstats.
2465 Patch provided by Heiko Schlittermann.
2467 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2468 Patch provided by Heiko Schlittermann.
2470 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2471 plus update to original patch.
2473 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2475 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2476 Patch provided by David Brownlee.
2478 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2480 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2482 NM/30 Bugzilla 888: TLS documentation bugfixes.
2484 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2486 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2487 Unlike the original bugzilla I have changed all shell scripts in src tree.
2489 NM/33 Bugzilla 898: Transport filter timeout fix.
2490 Patch by Todd Rinaldo.
2492 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2493 Patch by Serge Demonchaux.
2495 NM/35 Bugzilla 39: Base64 decode bug fixes.
2496 Patch by Jakob Hirsch.
2498 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2500 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2502 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2504 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2510 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2511 ALT-N's libdkim that I have put here:
2512 http://duncanthrax.net/exim-experimental/
2514 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2515 were called 'true' and it seems that C99 defines that as a reserved
2516 keyword to be used with 'bool' variable types. That means you could
2517 not include C99-style headers which use bools without triggering
2518 build errors in sieve.c.
2520 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2521 as mailq or other aliases. Changed the --help handling significantly
2522 to do whats expected. exim_usage() emits usage/help information.
2524 SC/01 Added the -bylocaldomain option to eximstats.
2526 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2528 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2530 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2536 PH/01 Another patch from the Sieve maintainer.
2538 PH/02 When an IPv6 address is converted to a string for single-key lookup
2539 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2540 dots are used instead of colons so that keys in lsearch files need not
2541 contain colons. This was done some time before quoting was made available
2542 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2543 (notated using the quote facility) so as to distinguish them from IPv4
2544 keys. This meant that lookups for IP addresses in host lists did not work
2545 for iplsearch lookups.
2547 This has been fixed by arranging for IPv6 addresses to be expressed with
2548 colons if the lookup type is iplsearch. This is not incompatible, because
2549 previously such lookups could never work.
2551 The situation is now rather anomalous, since one *can* have colons in
2552 ordinary lsearch keys. However, making the change in all cases is
2553 incompatible and would probably break a number of configurations.
2555 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2558 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2559 conversion specification without a maximum field width, thereby enabling
2560 a rogue spamd server to cause a buffer overflow. While nobody in their
2561 right mind would setup Exim to query an untrusted spamd server, an
2562 attacker that gains access to a server running spamd could potentially
2563 exploit this vulnerability to run arbitrary code as the Exim user.
2565 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2566 $primary_hostname instead of what libspf2 thinks the hosts name is.
2568 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2569 a directory entry by the name of the lookup key. Previously, if a
2570 symlink pointed to a non-existing file or a file in a directory that
2571 Exim lacked permissions to read, a lookup for a key matching that
2572 symlink would fail. Now it is enough that a matching directory entry
2573 exists, symlink or not. (Bugzilla 503.)
2575 PH/03 The body_linecount and body_zerocount variables are now exported in the
2578 PH/04 Added the $dnslist_matched variable.
2580 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2581 This means they are set thereafter only if the connection becomes
2584 PH/06 Added the client_condition to authenticators so that some can be skipped
2585 by clients under certain conditions.
2587 PH/07 The error message for a badly-placed control=no_multiline_responses left
2588 "_responses" off the end of the name.
2590 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2592 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2593 (without spaces) instead of just copying the configuration text.
2595 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2597 PH/11 Added $max_received_linelength.
2599 PH/12 Added +ignore_defer and +include_defer to host lists.
2601 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2602 way in which PCRE > 7.0 is built.
2604 PH/14 Implemented queue_only_load_latch.
2606 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2607 MAIL command. The effect was to mangle the value on 64-bit systems.
2609 PH/16 Another patch from the Sieve maintainer.
2611 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2613 PH/18 If a system quota error occurred while trying to create the file for
2614 a maildir delivery, the message "Mailbox is full" was not appended to the
2615 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2616 only to a quota excession during the actual writing of the file.
2618 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2619 characters?) which causes problems in log lines. The DN values are now
2620 passed through string_printing() before being added to log lines.
2622 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2623 and InterBase are left for another time.)
2625 PH/21 Added message_body_newlines option.
2627 PH/22 Guard against possible overflow in moan_check_errorcopy().
2629 PH/23 POSIX allows open() to be a macro; guard against that.
2631 PH/24 If the recipient of an error message contained an @ in the local part
2632 (suitably quoted, of course), incorrect values were put in $domain and
2633 $local_part during the evaluation of errors_copy.
2639 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2640 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2643 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2644 issue a MAIL command.
2646 PH/02 In an ACL statement such as
2648 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2650 if a client was not listed at all, or was listed with a value other than
2651 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2652 the condition was not true (as it should be), so access was not denied.
2653 The bug was that the ! inversion was incorrectly passed on to the second
2654 item. This has been fixed.
2656 PH/03 Added additional dnslists conditions == and =& which are different from
2657 = and & when the dns lookup returns more than one IP address.
2659 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2660 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2662 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2663 FSYNC, which compiles an option called disable_fsync that allows for
2664 bypassing fsync(). The documentation is heavily laced with warnings.
2666 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2668 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2669 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2670 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2671 including adding "make clean"; (3) Added -fPIC when compiling the test
2672 dynamically loaded module, to get rid of a warning.
2674 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2675 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2676 = 0s. The bug is otherwise harmless.
2678 PH/07 There was a bug in the dovecot authenticator such that the value of
2679 $auth1 could be overwritten, and so not correctly preserved, after a
2680 successful authentication. This usually meant that the value preserved by
2681 the server_setid option was incorrect.
2683 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2685 PH/09 Installed PCRE release 7.0.
2687 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2688 run for batched SMTP input. It is now run at the start of every message
2689 in the batch. While fixing this I discovered that the process information
2690 (output by running exiwhat) was not always getting set for -bs and -bS
2691 input. This is fixed, and it now also says "batched" for BSMTP.
2693 PH/11 Added control=no_pipelining.
2695 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2696 patch, slightly modified), and move the expansion of helo_data till after
2697 the connection is made in the smtp transport (so it can use these
2700 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2702 PH/14 Added log_selector = +pid.
2704 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2706 PH/16 Add ${if forany and ${if forall.
2708 PH/17 Added dsn_from option to vary the From: line in DSNs.
2710 PH/18 Flush SMTP output before performing a callout, unless control =
2711 no_callout_flush is set.
2713 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2714 was true (the default) a successful delivery failed to delete the retry
2715 item, thus causing premature timeout of the address. The bug is now
2718 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2720 PH/21 Long custom messages for fakedefer and fakereject are now split up
2721 into multiline responses in the same way that messages for "deny" and
2722 other ACL rejections are.
2724 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2725 with slight modification.
2727 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2728 draft, changing the syntax and factoring some duplicate code".
2730 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2731 for deliveries of the second and subsequent messages over the same SMTP
2734 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2735 ${reduce, with only minor "tidies".
2737 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2739 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2740 expansion side effects.
2742 PH/27 When a message times out after an over-quota error from an Exim-imposed
2743 quota, the bounce message says "mailbox is full". This message was not
2744 being given when it was a system quota that was exceeded. It now should
2747 MH/03 Made $recipients available in local_scan(). local_scan() already has
2748 better access to the recipient list through recipients_list[], but
2749 $recipients can be useful in postmaster-provided expansion strings.
2751 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2752 in the case of a MAIL command with additional options following the
2753 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2754 were accidentally chopped off.
2756 PH/29 SMTP synchronization checks are implemented when a command is read -
2757 there is a check that no more input is waiting when there shouldn't be
2758 any. However, for some commands, a delay in an ACL can mean that it is
2759 some time before the response is written. In this time, more input might
2760 arrive, invalidly. So now there are extra checks after an ACL has run for
2761 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2762 pipelining has not been advertised.
2764 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2766 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2767 returned in the SMTP response when smtp_return_error_details was false.
2768 This has been fixed.
2770 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2771 instead of the C I/O that was originally supplied, because problems were
2772 reported on Solaris.
2774 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2775 Exim which did not show up earlier: it was assuming that a call to
2776 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2777 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2778 was a macro that became an assignment, so it seemed to work. This has
2779 changed to a proper function call with a void return, hence the compile
2780 error. Exim's code has been fixed.
2782 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2785 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2787 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2789 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2790 because it turns out to be too big or there is a timeout) there is no
2791 "Completed" line in the log. When some messages of this type were
2792 selected by exigrep, they were listed as "not completed". Others were
2793 picked up by some special patterns. I have improved the selection
2794 criteria to be more general.
2796 PH/38 The host_find_failed option in the manualroute router can now be set
2797 to "ignore", to completely ignore a host whose IP address cannot be
2798 found. If all hosts are ignored, the behaviour is controlled by the new
2799 host_all_ignored option.
2801 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2802 homing or because of multiple MX records with /mx) generated more than
2803 one IP address, and the following item turned out to be the local host,
2804 all the secondary addresses of the first item were incorrectly removed
2805 from the list, along with the local host and any following hosts (which
2806 is what is supposed to happen).
2808 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2809 whoever called Exim into the -H file. In the case of the daemon it was
2810 behaving confusingly. When first started, it used values for whoever
2811 started the daemon, but after a SIGHUP it used the Exim user (because it
2812 calls itself on a restart). I have changed the code so that it now always
2815 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2816 message are rejected with the same error (e.g. no authentication or bad
2817 sender address), and a DATA command is nevertheless sent (as can happen
2818 with PIPELINING or a stupid MUA), the error message that was given to the
2819 RCPT commands is included in the rejection of the DATA command. This is
2820 intended to be helpful for MUAs that show only the final error to their
2823 PH/42 Another patch from the Sieve maintainer.
2825 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2826 Eximstats - Fixed some broken HTML links and added missing column headers
2828 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2829 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2831 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2833 PH/43 Yet another patch from the Sieve maintainer.
2835 PH/44 I found a way to check for a TCP/IP connection going away before sending
2836 the response to the final '.' that terminates a message, but only in the
2837 case where the client has not sent further data following the '.'
2838 (unfortunately, this is allowed). However, in many cases there won't be
2839 any further data because there won't be any more messages to send. A call
2840 to select() can be used: if it shows that the input is "ready", there is
2841 either input waiting, or the socket has been closed. An attempt to read
2842 the next input character can distinguish the two cases. Previously, Exim
2843 would have sent an OK response which the client would never have see.
2844 This could lead to message repetition. This fix should cure that, at
2845 least in a lot of common cases.
2847 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2848 advertised in response to EHLO.
2854 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2855 fixed by 4.65/MH/01 (is this a record?) are fixed:
2857 (i) An empty string was always treated as zero by the numeric comparison
2858 operators. This behaviour has been restored.
2860 (ii) It is documented that the numeric comparison operators always treat
2861 their arguments as decimal numbers. This was broken in that numbers
2862 starting with 0 were being interpreted as octal.
2864 While fixing these problems I realized that there was another issue that
2865 hadn't been noticed. Values of message_size_limit (both the global option
2866 and the transport option) were treated as octal if they started with 0.
2867 The documentation was vague. These values are now always treated as
2868 decimal, and I will make that clear in the documentation.
2874 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2875 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2878 MH/01 Don't check that the operands of numeric comparison operators are
2879 integers when their expansion is in "skipping" mode (fixes bug
2880 introduced by 4.64-PH/07).
2882 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2883 child addresses, Exim now panics and dies. Previously, because the count
2884 is held in a short int, deliveries were likely to be lost. As such a
2885 large number of recipients for a single message is ridiculous
2886 (performance will be very, very poor), I have chosen to impose a limit
2887 rather than extend the field.
2893 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2894 leftover -K file (the existence of which was triggered by #402).
2895 While we were at it, introduced process PID as part of the -K
2896 filename. This should rule out race conditions when creating
2899 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2900 processing considerably. Previous code took too long for large mails,
2901 triggering a timeout which in turn triggers #401.
2903 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2904 in the DK code in transports.c. sendfile() is not really portable,
2905 hence the _LINUX specificness.
2907 TF/01 In the add_headers option to the mail command in an Exim filter,
2908 there was a bug that Exim would claim a syntax error in any
2909 header after the first one which had an odd number of characters
2912 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2913 callout verification, Exim cached a "reject" for the entire domain. This
2914 is correct for most verifications, but it is not correct for a recipient
2915 verification with use_sender or use_postmaster set, because in that case
2916 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2917 case of MAIL FROM:<> rejection from other early rejections (e.g.
2918 rejection of HELO). When verifying a recipient using a non-null MAIL
2919 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2920 Whatever the result of the callout, the value of the domain cache is
2921 left unchanged (for any other kind of callout, getting as far as trying
2922 RCPT means that the domain itself is ok).
2924 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2927 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2928 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2929 session, the daemon ignores ECONNECT errors and logs others; it now
2930 ignores EPIPE as well.
2932 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2933 (quoted-printable decoding).
2935 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2936 later the small subsequent patch to fix an introduced bug.
2938 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2940 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2942 PH/08 An error is now given if message_size_limit is specified negative.
2944 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2945 to be given (somewhat) arbitrary names.
2947 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2950 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2951 miscellaneous code fixes
2953 PH/10 Added the log_reject_target ACL modifier to specify where to log
2956 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2957 hostname. This is wrong, because it relates to the incoming message (and
2958 probably the interface on which it is arriving) and not to the outgoing
2959 callout (which could be using a different interface). This has been
2960 changed to use the value of the helo_data option from the smtp transport
2961 instead - this is what is used when a message is actually being sent. If
2962 there is no remote transport (possible with a router that sets up host
2963 addresses), $smtp_active_hostname is used.
2965 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2966 tweaks were necessary in order to get it to work (see also 21 below):
2967 (a) The code assumed that strncpy() returns a negative number on buffer
2968 overflow, which isn't the case. Replaced with Exim's string_format()
2970 (b) There were several signed/unsigned issues. I just did the minimum
2971 hacking in of casts. There is scope for a larger refactoring.
2972 (c) The code used strcasecmp() which is not a standard C function.
2973 Replaced with Exim's strcmpic() function.
2974 (d) The code set only $1; it now sets $auth1 as well.
2975 (e) A simple test gave the error "authentication client didn't specify
2976 service in request". It would seem that Dovecot has changed its
2977 interface. Fortunately there's a specification; I followed it and
2978 changed what the client sends and it appears to be working now.
2980 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2983 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2984 address A is aliased to B and C, where B exists and C does not. Without
2985 -v the output is "A verified" because verification stops after a
2986 successful redirection if more than one address is generated. However,
2987 with -v the child addresses are also verified. Exim was outputting "A
2988 failed to verify" and then showing the successful verification for C,
2989 with its parentage. It now outputs "B failed to verify", showing B's
2990 parentage before showing the successful verification of C.
2992 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2993 look up a TXT record in a specific list after matching in a combined
2996 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2997 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2998 they consult the DNS. I had assumed they would set it the way they
2999 wanted; and indeed my experiments on Linux seem to show that in some
3000 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3001 To be on the safe side, however, I have now made the interface to
3002 host_find_byname() similar to host_find_bydns(), with an argument
3003 containing the DNS resolver options. The host_find_byname() function now
3004 sets these options at its start, just as host_find_bydns() does. The smtp
3005 transport options dns_qualify_single and dns_search_parents are passed to
3006 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3007 of host_find_byname() use the default settings of RES_DEFNAMES
3008 (qualify_single) but not RES_DNSRCH (search_parents).
3010 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3011 spool_read_header() do less string testing, by means of a preliminary
3012 switch on the second character of optional "-foo" lines. (This is
3013 overdue, caused by the large number of possibilities that now exist.
3014 Originally there were few.) While I was there, I also converted the
3015 str(n)cmp tests so they don't re-test the leading "-" and the first
3016 character, in the hope this might squeeze out yet more improvement.
3018 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3019 flag allowing group syntax was set by the header_syntax check but not
3020 turned off, possible causing trouble later; (2) The flag was not being
3021 set at all for the header_verify test, causing "group"-style headers to
3022 be rejected. I have now set it in this case, and also caused header_
3023 verify to ignore an empty address taken from a group. While doing this, I
3024 came across some other cases where the code for allowing group syntax
3025 while scanning a header line wasn't quite right (mostly, not resetting
3026 the flag correctly in the right place). These bugs could have caused
3027 trouble for malformed header lines. I hope it is now all correct.
3029 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3030 with the "reply" argument non-NULL. The code, however (which originally
3031 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3032 but it didn't always do it. This confused somebody who was copying the
3033 code for some other use. I have removed all the tests.
3035 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3036 feature that was used to support insecure browsers during the U.S. crypto
3037 embargo. It requires special client support, and Exim is probably the
3038 only MTA that supported it -- and would never use it because real RSA is
3039 always available. This code has been removed, because it had the bad
3040 effect of slowing Exim down by computing (never used) parameters for the
3041 RSA_EXPORT functionality.
3043 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3044 authenticator to fail if there's a tab character in the incoming data
3045 (there should never be unless someone is messing about, as it's supposed
3046 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3047 if the connection is using TLS or if the remote IP is the same as the
3048 local IP, and the "valid-client-cert option" if a client certificate has
3051 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3052 authenticators. This can be used for authorization after authentication
3053 succeeds. (In the case of plaintext, it servers for both authentication
3056 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3057 if any retry times were supplied.
3059 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3060 connection, where there is no client IP address to check. In this
3061 situation, the verify now always succeeds.
3063 PH/25 Applied John Jetmore's -Mset patch.
3065 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3067 PH/27 In a string expansion for a processed (not raw) header when multiple
3068 headers of the same name were present, leading whitespace was being
3069 removed from all of them, but trailing whitespace was being removed only
3070 from the last one. Now trailing whitespace is removed from each header
3071 before concatenation. Completely empty headers in a concatenation (as
3072 before) are ignored.
3074 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3075 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3077 PH/29 [Removed. This was a change that I later backed out, and forgot to
3078 correct the ChangeLog entry (that I had efficiently created) before
3079 committing the later change.]
3081 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3082 address errors (4xx response to RCPT) over the same connection as other
3083 messages routed to the same hosts. Such deliveries are always "forced",
3084 so retry times are not inspected. This resulted in far too many retries
3085 for the affected addresses. The effect occurred only when there were more
3086 hosts than the hosts_max_try setting in the smtp transport when it had
3087 the 4xx errors. Those hosts that it had tried were not added to the list
3088 of hosts for which the message was waiting, so if all were tried, there
3089 was no problem. Two fixes have been applied:
3091 (i) If there are any address or message errors in an SMTP delivery, none
3092 of the hosts (tried or untried) are now added to the list of hosts
3093 for which the message is waiting, so the message should not be a
3094 candidate for sending over the same connection that was used for a
3095 successful delivery of some other message. This seems entirely
3096 reasonable: after all the message is NOT "waiting for some host".
3097 This is so "obvious" that I'm not sure why it wasn't done
3098 previously. Hope I haven't missed anything, but it can't do any
3099 harm, as the worst effect is to miss an optimization.
3101 (ii) If, despite (i), such a delivery is accidentally attempted, the
3102 routing retry time is respected, so at least it doesn't keep
3103 hammering the server.
3105 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3106 in ${readsocket because some servers need this prod.
3108 PH/32 Added some extra debug output when updating a wait-xxx database.
3110 PH/33 The hint "could be header name not terminated by colon", which has been
3111 given for certain expansion errors for a long time, was not being given
3112 for the ${if def:h_colon_omitted{... case.
3114 PH/34 The spec says: "With one important exception, whenever a domain list is
3115 being scanned, $domain contains the subject domain." There was at least
3116 one case where this was not true.
3118 PH/35 The error "getsockname() failed: connection reset by peer" was being
3119 written to the panic log as well as the main log, but it isn't really
3120 panic-worthy as it just means the connection died rather early on. I have
3121 removed the panic log writing for the ECONNRESET error when getsockname()
3124 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3125 runs only) independently of the message's sender address. This meant
3126 that, if the 4xx error was in fact related to the sender, a different
3127 message to the same recipient with a different sender could confuse
3128 things. In particular, this can happen when sending to a greylisting
3129 server, but other circumstances could also provoke similar problems.
3130 I have changed the default so that the retry time for these errors is now
3131 based a combination of the sender and recipient addresses. This change
3132 can be overridden by setting address_retry_include_sender=false in the
3135 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3136 remote server are returned as part of bounce messages. This was not
3137 happening for LMTP over a pipe (the lmtp transport), but now it is the
3138 same for both kinds of LMTP.
3140 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3141 in header lines that were in fact CNAMEs. This is no longer the case.
3143 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3144 and queue runs started by the daemon processed all messages. This has
3145 been fixed so that -R and -S can now usefully be given with -q<time>.
3147 PH/40 Import PCRE release 6.7 (fixes some bugs).
3149 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3151 PH/42 Give an error if -q is specified more than once.
3153 PH/43 Renamed the variables $interface_address and $interface_port as
3154 $received_ip_address and $received_port, to make it clear that these
3155 values apply to message reception, and not to the outgoing interface when
3156 a message is delivered. (The old names remain recognized, of course.)
3158 PH/44 There was no timeout on the connect() call when using a Unix domain
3159 socket in the ${readsocket expansion. There now is.
3161 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3162 be meaningful with "accept".
3164 SC/01 Eximstats V1.43
3165 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3167 SC/02 Eximstats V1.44
3168 Use a glob alias rather than an array ref in the generated
3169 parser. This improves both readability and performance.
3171 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3172 Collect SpamAssassin and rejection statistics.
3173 Don't display local sender or destination tables unless
3174 there is data to show.
3175 Added average volumes into the top table text output.
3177 SC/04 Eximstats V1.46
3178 Collect data on the number of addresses (recipients)
3179 as well as the number of messages.
3181 SC/05 Eximstats V1.47
3182 Added 'Message too big' to the list of mail rejection
3183 reasons (thanks to Marco Gaiarin).
3185 SC/06 Eximstats V1.48
3186 Mainlog lines which have GMT offsets and are too short to
3187 have a flag are now skipped.
3189 SC/07 Eximstats V1.49 (Alain Williams)
3190 Added the -emptyok flag.
3192 SC/08 Eximstats V1.50
3193 Fixes for obtaining the IP address from reject messages.
3195 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3196 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3197 whitespace changes from 4.64-PH/27
3199 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3202 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3203 are found, allow negative numbers in numeric criteria)
3205 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3207 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3208 to match changes made in 4.64-PH/43
3210 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3212 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3213 "domain" is actually an IP address, and force a failure. This locks out
3214 those revolvers/nameservers that support "A-for-A" lookups, in
3215 contravention of the specifications.
3217 PH/48 When a host name was looked up from an IP address, and the subsequent
3218 forward lookup of the name timed out, the host name was left in
3219 $sender_host_name, contrary to the specification.
3221 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3222 restricted to single-key lookups, Exim was not diagnosing an error if
3223 * or *@ was used with a query-style lookup.
3225 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3227 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3228 long ago, but noone interested enough thought of it. Let's just say that
3229 the "1.1" means that there are some new functions that weren't there at
3230 some point in the past.
3232 PH/51 Error processing for expansion failure of helo_data from an smtp
3233 transport during callout processing was broken.
3235 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3236 tested/used via the -bh/-bhc/-bs options.
3238 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3239 bug, fixed in subsequent PCRE releases).
3241 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3242 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3244 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3250 SC/01 Use a glob alias rather than an array ref in eximstats generated
3251 parser. This improves both readability and performance.
3253 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3254 Don't display local sender or destination tables in eximstats unless
3255 there is data to show.
3256 Added average volumes into the eximstats top table text output.
3258 SC/03 Collect data on the number of addresses (recipients) as well
3259 as the number of messages in eximstats.
3261 TF/01 Correct an error in the documentation for the redirect router. Exim
3262 does (usually) call initgroups() when daemonizing.
3264 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3265 with consistent privilege compared to when running as a daemon.
3267 TF/03 Note in the spec that $authenticated_id is not set for local
3268 submissions from trusted users.
3270 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3271 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3273 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3274 by adding some example configuration directives to the default
3275 configuration file. A little bit of work is required to uncomment the
3276 directives and define how usernames and passwords are checked, but
3277 there is now a framework to start from.
3279 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3280 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3281 without this. I don't know how relevant this is to other LDAP libraries.
3283 PH/02 Add the verb name to the "unknown ACL verb" error.
3285 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3287 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3289 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3290 directory not expanded when it should be if an expanded home directory
3291 was set for the address (which is overridden by the transport).
3293 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3296 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3297 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3298 because it is too late at that time, and has no effect.
3300 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3301 security issue with \' (bugzilla #107). I could not use the
3302 PQescapeStringConn() function, because it needs a PGconn value as one of
3305 PH/08 When testing addresses using -bt, indicate those final addresses that
3306 are duplicates that would not cause an additional delivery. At least one
3307 person was confused, thinking that -bt output corresponded to deliveries.
3308 (Suppressing duplicates isn't a good idea as you lose the information
3309 about possibly different redirections that led to the duplicates.)
3311 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3312 systems where poll() doesn't work, in particular OS X.
3314 PH/10 Added more information to debugging output for retry time not reached.
3316 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3317 operations in malware.c.
3319 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3322 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3323 syslog_duplication set false, log lines that would normally be written
3324 both the the main log and to the reject log were not written to syslog at
3327 PH/14 In the default configuration, change the use of "message" in ACL warn
3328 statements to "add_header".
3330 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3331 not followed by a command (e.g. "seen endif").
3333 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3334 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3337 PH/17 Added extra conditions to the default value of delay_warning_condition
3341 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3342 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3343 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3346 The Auto-Submitted: and various List- headers are standardised, whereas I
3347 don't think Precedence: ever was.
3349 PH/18 Refactored debugging code in route_finduser() to show more information,
3350 in particular, the error code if getpwnam() issues one.
3352 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3353 This is apparently needed in addition to the PH/07 change above to avoid
3354 any possible encoding problems.
3356 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3357 but not after initializing Perl.
3359 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3360 output them only if debugging. By default they are written stderr,
3361 apparently, which is not desirable.
3363 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3366 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3369 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3371 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3372 authenticated or an ident call has been made. Suppress the default
3373 values for $authenticated_id and $authenticated_sender (but permit -oMai
3374 and -oMas) when testing with -bh.
3376 PH/24 Re-jigged the order of the tests in the default configuration so that the
3377 tests for valid domains and recipients precede the DNS black list and CSA
3378 tests, on the grounds that those ones are more expensive.
3380 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3381 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3382 This bug exists in every version of Exim that I still have, right back to
3385 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3386 However, an attempt to turn on case-sensitivity in a regex key by
3387 including (?-i) didn't work because the subject string was already
3388 lowercased, and the effects were non-intuitive. It turns out that a
3389 one-line patch can be used to allow (?-i) to work as expected.
3395 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3396 other effects) broke the use of negated acl sub-conditions.
3398 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3401 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3402 "Deny" causes Exim to reject the incoming connection with a 554 error.
3403 Unfortunately, if there is a major crisis, such as a disk failure,
3404 tcp-wrappers gives "deny", whereas what one would like would be some
3405 kind of temporary error. A kludge has been added to help with this.
3406 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3407 554 error is used if errno is still zero or contains ENOENT (which occurs
3408 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3411 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3413 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3414 errors. Otherwise a message that provokes a temporary error (when other
3415 messages do not) can cause a whole host to time out.
3417 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3418 addresses were routed directly to files or pipes from a redirect router.
3419 File deliveries just didn't batch; pipe deliveries might have suffered
3422 PH/06 A failure to get a lock for a hints database would erroneously always say
3423 "Failed to get write lock", even when it was really a read lock.
3425 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3426 of 0600. This has been changed to use the value of the lockfile_mode
3427 option (which defaults to 0600).
3429 PH/08 Applied small patch from the Sieve maintainer.
3431 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3432 folder from quota calculations, a direct delivery into this folder messed
3433 up the contents of the maildirsize file. This was because the regex was
3434 used only to exclude .Trash (or whatever) when the size of the mailbox
3435 was calculated. There was no check that a delivery was happening into an
3436 excluded directory. This bug has been fixed by ignoring all quota
3437 processing for deliveries into excluded directories.
3439 PH/10 Added the maildirfolder_create_regex option to appendfile.
3445 PH/01 The code for finding all the local interface addresses on a FreeBSD
3446 system running IPv6 was broken. This may well have applied to all BSD
3447 systems, as well as to others that have similar system calls. The broken
3448 code found IPv4 interfaces correctly, but gave incorrect values for the
3449 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3450 that it would not match correctly against @[] and not recognize the IPv6
3453 PH/02 The ipliteral router was not recognizing addresses of the form user@
3454 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3456 PH/03 Added disable_ipv6.
3458 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3459 decoded form, because it is most often used to construct To: headers
3460 lines in autoreplies, and the decoded form may well be syntactically
3461 invalid. However, $reply_address has leading white space removed, and all
3462 newlines turned into spaces so that the autoreply transport does not
3465 PH/05 If group was specified without a user on a router, and no group or user
3466 was specified on a transport, the group from the router was ignored.
3468 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3469 for visible compile-time settings that can be used to change these
3470 numbers, for those that want even more. Backwards compatibility with old
3471 spool files has been maintained. However, going back to a previous Exim
3472 release will lost any variables that are in spool files.
3474 PH/07 Two small changes when running in the test harness: increase delay when
3475 passing a TCP/IP connection to a new process, in case the original
3476 process has to generate a bounce, and remove special handling of
3477 127.0.0.2 (sic), which is no longer necessary.
3479 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3480 be the same on different OS.
3482 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3485 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3486 whether --show-vars was specified or not
3488 JJ/02 exipick: Added support for new ACL variable spool format introduced
3491 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3492 syntactically invalid From: or Reply-to: line, and a filter used this to
3493 generate an autoreply, and therefore failed to obtain an address for the
3494 autoreply, Exim could try to deliver to a non-existent relative file
3495 name, causing unrelated and misleading errors. What now happens is that
3496 it logs this as a hard delivery error, but does not attempt to create a
3499 PH/11 The exinext utility has a -C option for testing purposes, but although
3500 the given file was scanned by exinext itself; it wasn't being passed on
3501 when Exim was called.
3503 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3504 an end-of-file indication when reading a command response.
3506 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3507 compiled. In many other places in Exim, IPv6 addresses are always
3508 recognized, so I have changed this. It also means that IPv4 domain
3509 literals of the form [IPV4:n.n.n.n] are now always recognized.
3511 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3512 used if the router is not running as root, for example, when verifying at
3513 ACL time, or when using -bh. The debugging output from this situation was
3514 non-existent - all you got was a failure to exec. I have made two
3517 (a) Failures to set uid/gid, the current directory, or a process leader
3518 in a subprocess such as that created by queryprogram now generate
3519 suitable debugging output when -d is set.
3521 (b) The queryprogram router detects when it is not running as root,
3522 outputs suitable debugging information if -d is set, and then runs
3523 the subprocess without attempting to change uid/gid.
3525 PH/15 Minor change to Makefile for building test_host (undocumented testing
3528 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3529 additional section of a DNS packet that returns MX or SRV records.
3530 Instead, it always explicitly searches for A/AAAA records. This avoids
3531 major problems that occur when a DNS server includes only records of one
3532 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3533 fixed another bug: if SRV records were looked up and the corresponding
3534 address records were *not* found in the additional section, the port
3535 values from the SRV records were lost.
3537 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3538 using the correct key (the original address) when searching the retry
3539 rules in order to find which one to use for generating the retry hint.
3541 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3542 adding the default one. Similarly, if it contains a Reply-To: header, the
3543 errors_reply_to option, if set, is not used.
3545 PH/19 When calculating a retry time, Exim used to measure the "time since
3546 failure" by looking at the "first failed" field in the retry record. Now
3547 it does not use this if it is later than than the arrival time of the
3548 message. Instead it uses the arrival time. This makes for better
3549 behaviour in cases where some deliveries succeed, thus re-setting the
3550 "first failed" field. An example is a quota failure for a huge message
3551 when small messages continue to be delivered. Without this change, the
3552 "time since failure" will always be short, possible causing more frequent
3553 delivery attempts for the huge message than are intended.
3554 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3556 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3557 $1, $2, $3) because the numerical variables can be reset during some
3558 expansion items (e.g. "match"), thereby losing the authentication data.
3560 PH/21 Make -bV show the size of off_t variables so that the test suite can
3561 decide whether to run tests for quotas > 2G.
3563 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3564 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3565 filecount value is greater than 2G or if a quota value is greater than 2G
3566 on a system where the size of off_t is not greater than 4, a panic error
3569 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3570 never match. The debug and -bh output now contains an explicit error
3571 message indicating a malformed IPv4 address or mask.
3573 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3574 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3575 PH/23 above applies.
3577 PH/25 Do not write to syslog when running in the test harness. The only
3578 occasion when this arises is a failure to open the main or panic logs
3579 (for which there is an explicit test).
3581 PH/26 Added the /no_tell option to "control=freeze".
3583 PH/27 If a host name lookup failed very early in a connection, for example, if
3584 the IP address matched host_lookup and the reverse lookup yielded a name
3585 that did not have a forward lookup, an error message of the form "no IP
3586 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3587 could be logged. Now it outputs the IP address instead of "NULL".
3589 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3590 allows the sender and the authenticated sender to be set when
3591 submitting a message from within Exim. Since child_open_exim() is
3592 documented for local_scan(), the new function should be too.
3594 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3595 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3596 results in an empty string is now treated as unset.
3598 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3600 PH/31 Added sender_verify_fail logging option.
3602 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3603 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3604 tidied the source and removed it altogether.
3606 PH/33 When a queue run was abandoned because the load average was too high, a
3607 log line was always written; now it is written only if the queue_run log
3608 selector is set. In addition, the log line for abandonment now contains
3609 information about the queue run such as the pid. This is always present
3610 in "start" and "stop" lines but was omitted from the "abandon" line.
3612 PH/34 Omit spaces between a header name and the colon in the error message that
3613 is given when verify = headers_syntax fails (if there are lots of them,
3614 the message gets confusing).
3616 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3617 names, as there are now some PTR records that contain slashes. This check
3618 is only to protect against broken name servers that fall over on strange
3619 characters, so the fact that it applies to all lookups doesn't matter.
3621 PH/36 Now that the new test suite is complete, we can remove some of the
3622 special code in Exim that was needed for the old test suite. For example,
3623 sorting DNS records because real resolvers return them in an arbitrary
3624 order. The new test suite's fake resolver always returns records in the
3627 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3628 bounces) except when queue_only is set, to avoid logging races between
3629 the different processes.
3631 PH/38 Panic-die if .include specifies a non-absolute path.
3633 PH/39 A tweak to the "H" retry rule from its user.
3635 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3636 a label. They prevented compilation on older perls.
3638 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3639 a warning to be raised on newish perls.
3641 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3642 on queue. Changes to match documented behaviour of showing count of
3643 messages matching specified criteria.
3645 PH/40 Changed the default ident timeout from 30s to 5s.
3647 PH/41 Added support for the use of login_cap features, on those BSD systems
3648 that have them, for controlling the resources used by pipe deliveries.
3650 PH/42 The content-scanning code uses fopen() to create files in which to put
3651 message data. Previously it was not paying any attention to the mode of
3652 the files. Exim runs with umask(0) because the rest of the code creates
3653 files with open(), and sets the required mode explicitly. Thus, these
3654 files were ending up world-writeable. This was not a big issue, because,
3655 being within the spool directory, they were not world-accessible. I have
3656 created a function called modefopen, which takes an additional mode
3657 argument. It sets umask(777), creates the file, chmods it to the required
3658 mode, then resets the umask. All the relevant calls to fopen() in the
3659 content scanning code have been changed to use this function.
3661 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3662 to 24 hours. This avoids potential overflow problems when processing G
3663 and H retry rules. I suspect nobody ever tinkers with this value.
3665 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3667 PH/45 When the plaintext authenticator is running as a client, the server's
3668 challenges are checked to ensure they are valid base64 strings. By
3669 default, the authentication attempt is cancelled if an invalid string is
3670 received. Setting client_ignore_invalid_base64 true ignores these errors.
3671 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3672 they are received. Thus, the responses can be made to depend on the
3673 challenges. If an invalid string is ignored, an empty string is placed in
3676 PH/46 Messages that are created by the autoreply transport now contains a
3677 References: header, in accordance with RFCs 2822 and 3834.
3679 PH/47 Added authenticated_sender_force to the smtp transport.
3681 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3683 PH/49 Installed latest patch from the Sieve maintainer.
3685 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3686 was also set, the appendfile transport was unnecessarily scanning a
3687 directory of message files (e.g. for maildir delivery) to find the count
3688 of files (along with the size), even though it did not need this
3689 information. It now does the scan only if it needs to find either the
3690 size of the count of files.
3692 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3694 PH/52 Two bugs concerned with error handling when the smtp transport is
3697 (i) Exim was not creating retry information for temporary errors given
3698 for individual recipients after the DATA command when the smtp transport
3699 was used in LMTP mode. This meant that they could be retried too
3700 frequently, and not timed out correctly.
3702 (ii) Exim was setting the flag that allows error details to be returned
3703 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3704 recipients that were returned after the DATA command.
3706 PH/53 This is related to PH/52, but is more general: for any failing address,
3707 when detailed error information was permitted to be returned to the
3708 sender, but the error was temporary, then after the final timeout, only
3709 "retry timeout exceeded" was returned. Now it returns the full error as
3710 well as "retry timeout exceeded".
3712 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3713 do this, and (what is worse) MTAs that accept it.
3715 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3716 will now be deprecated.
3718 PH/56 New os.c-cygwin from the Cygwin maintainer.
3720 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3721 formats (previously only available in exim formats via -bpr, -bpru,
3722 and -bpra. Now also available in native and exiqgrep formats)
3724 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3725 with very large, slow to parse queues
3727 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3729 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3731 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3732 responses to authentication challenges, though it was showing the
3733 challenges; (ii) I've removed the CR characters from the debug output for
3736 PH/58 Allow for the insertion of a newline as well as a space when a string
3737 is turned into more than one encoded-word during RFC 2047 encoding. The
3738 Sieve code now uses this.
3740 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3741 data_4xx, lost_connection, tls_required.
3743 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3744 message was being sent as an SMTP response.
3746 PH/61 Add -l and -k options to exicyclog.
3748 PH/62 When verifying, if an address was redirected to one new address, so that
3749 verification continued, and the new address failed or deferred after
3750 having set something in $address_data, the value of $address_data was not
3751 passed back to the ACL. This was different to the case when no
3752 redirection occurred. The value is now passed back in both cases.
3754 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3755 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3756 use login_cap.h, so on its own it isn't the distinguishing feature. The
3757 new name refers directly to the setclassresources() function.
3759 PH/65 Added configuration files for NetBSD3.
3761 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3763 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3764 is preferred over IPv4.
3766 PH/68 The bounce_return_message and bounce_return_body options were not being
3767 honoured for bounces generated during the reception of non-SMTP messages.
3768 In particular, this applied to messages rejected by the ACL. This bug has
3769 been fixed. However, if bounce_return_message is true and bounce_return_
3770 body is false, the headers that are returned for a non-SMTP message
3771 include only those that have been read before the error was detected.
3772 (In the case of an ACL rejection, they have all been read.)
3774 PH/69 The HTML version of the specification is now built in a directory called
3775 spec_html instead of spec.html, because the latter looks like a path with
3776 a MIME-type, and this confuses some software.
3778 PH/70 Catch two compiler warnings in sieve.c.
3780 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3781 function verify_get_ident() calls ip_connect() to connect a socket, but
3782 if the "connect()" function timed out, ip_connect() used to close the
3783 socket. However, verify_get_ident() also closes the socket later, and in
3784 between Exim writes to the log, which may get opened at this point. When
3785 the socket was closed in ip_connect(), the log could get the same file
3786 descriptor number as the socket. This naturally causes chaos. The fix is
3787 not to close the socket in ip_connect(); the socket should be closed by
3788 the function that creates it. There was only one place in the code where
3789 this was missing, in the iplookup router, which I don't think anybody now
3790 uses, but I've fixed it anyway.
3792 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3793 well as to direct DNS lookups. Otherwise the handling of names in host
3794 lists is inconsistent and therefore confusing.
3800 PH/01 Two changes to the default runtime configuration:
3802 (1) Move the checks for relay_from_hosts and authenticated clients from
3803 after to before the (commented out) DNS black list checks.
3805 (2) Add control=submission to the relay_from_hosts and authenticated
3806 clients checks, on the grounds that messages accepted by these
3807 statements are most likely to be submissions.
3809 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3811 (1) Generate an error if the third argument for the ${prvs expansion is
3814 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3817 (3) Reset the variables that are obtained from the first argument of
3818 ${prvscheck and used in the second argument before leaving the code,
3819 because their memory is reclaimed, so using them afterwards may do
3822 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3823 one (it's much easier than Tom thought :-).
3825 (5) Because of (4), we can now allow for the use of $prvscheck_result
3826 inside the third argument.
3828 PH/03 For some reason, the default setting of PATH when running a command from
3829 a pipe transport was just "/usr/bin". I have changed it to
3832 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3833 anything to be listed in the output from -bV.
3835 PH/05 When a filter generated an autoreply, the entire To: header line was
3836 quoted in the delivery log line, like this:
3838 => >A.N.Other <ano@some.domain> <original@ddress> ...
3840 This has been changed so that it extracts the operative address. There
3841 may be more than one such address. If so, they are comma-separated, like
3844 => >ano@some.domain,ona@other.domain <original@ddress> ...
3846 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3847 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3848 not being looked up in the rDNS to get a host name, Exim was showing the
3849 IP address twice in Received: lines, even though the IP addresses were
3850 identical. For example:
3852 Received: from [1.2.3.4] (helo=[1.2.3.4])
3854 However, if the real host name was known, it was omitting the HELO data
3855 if it matched the actual IP address. This has been tidied up so that it
3856 doesn't show the same IP address twice.
3858 PH/07 When both +timestamp and +memory debugging was on, the value given by
3859 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3860 called by the debug printing, thereby overwriting the timestamp buffer.
3861 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3863 PH/08 When the original message was included in an autoreply transport, it
3864 always said "this is a copy of the message, including all the headers",
3865 even if body_only or headers_only was set. It now gives an appropriate
3868 PH/09 Applied a patch from the Sieve maintainer which:
3870 o fixes some comments
3871 o adds the (disabled) notify extension core
3872 o adds some debug output for the result of if/elsif tests
3873 o points to the current vacation draft in the documentation
3874 and documents the missing references header update
3878 o fixes a bug in processing the envelope test (when testing
3879 multiple envelope elements, the last element determined the
3882 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3883 Electronic Mail") by including:
3885 Auto-submitted: auto-generated
3887 in the messages that it generates (bounce messages and others, such as
3888 warnings). In the case of bounce messages for non-SMTP messages, there was
3889 also a typo: it was using "Auto_submitted" (underscore instead of
3890 hyphen). Since every message generated by Exim is necessarily in response
3891 to another message, thes have all been changed to:
3893 Auto-Submitted: auto-replied
3895 in accordance with these statements in the RFC:
3897 The auto-replied keyword:
3899 - SHOULD be used on messages sent in direct response to another
3900 message by an automatic process,
3902 - MUST NOT be used on manually-generated messages,
3904 - MAY be used on Delivery Status Notifications (DSNs) and Message
3905 Disposition Notifications (MDNs),
3907 - MUST NOT be used on messages generated by automatic or periodic
3908 processes, except for messages which are automatic responses to
3911 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3912 to the default Received: header definition.
3914 PH/12 Added log selector acl_warn_skipped (default on).
3916 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3917 variables because (a) they are in the wrong storage pool and (b) even if
3918 they were copied, it wouldn't work properly because of the caching.
3920 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3921 checking when decoding. Apparently there are clients that generate
3922 overlong encoded strings. Why am I not surprised?
3924 PH/15 If the first argument of "${if match_address" was not empty, but did not
3925 contain an "@" character, Exim crashed. Now it writes a panic log message
3926 and treats the condition as false.
3928 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3930 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3931 extension "envelope-auth". The code is finished and in agreement with
3932 other implementations, but there is no documentation so far and in fact,
3933 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3934 not changing the active code.
3936 Print executed "if" and "elsif" statements when debugging is used. This
3937 helps a great deal to understand what a filter does.
3939 Document more things not specified clearly in RFC3028. I had all this
3940 sorted out, when out of a sudden new issues came to my mind. Oops."
3942 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3945 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3946 canonical form (as documented). However, after a host name lookup from
3947 the IP address, check_host() was doing a simple string comparison with
3948 addresses acquired from the DNS when checking that the found name did
3949 have the original IP as one of its addresses. Since any found IPv6
3950 addresses are likely to be in abbreviated form, the comparison could
3951 fail. Luckily, there already exists a function for doing the comparison
3952 by converting both addresses to binary, so now that is used instead of
3953 the text comparison.
3955 PH/20 There was another similar case to PH/19, when a complete host name was
3956 given in a host list; looking up its IP address could give an abbreviated
3957 form, whereas the current host's name might or might not be abbreviated.
3958 The same fix has been applied.
3964 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3965 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3968 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3969 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3971 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3973 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3974 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3975 Also, for "Auto-Submitted", treat anything other than "no" as
3976 non-personal, in accordance with RFC 3834. (Previously it treated
3977 anything starting "auto-" as non-personal.)
3979 TF/01 The control=submission/name=... option had a problem with syntax
3980 errors if the name included a slash character. The /name= option
3981 now slurps the rest of the string, so it can include any characters
3982 but it must come last in the list of options (after /sender_retain
3985 PH/05 Some modifications to the interface to the fake nameserver for the new
3993 TK/01 Added the "success_on_redirect" address verification option. See
3994 NewStuff for rationale and an example.
3996 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3998 PH/02 Patch to exigrep to allow it to work on syslog lines.
4000 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4001 fread() to skip over the body file's header line, because in Cygwin the
4002 header line is locked and is inaccessible.
4004 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4005 co-exist for some time) to make it clear that it is the Exim ID that is
4006 referenced, not the Message-ID: header line.
4008 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4009 string_format() function, because snprintf() does not exist on all
4012 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4013 ${stat: expansion item.
4015 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4016 protocol synchronization error", to keep the pedants happy.
4018 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4019 well as for IRIX systems, when gcc is being used. See the host.c source
4022 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4024 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4027 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4028 transport and to the smtp transport in LMTP mode.
4030 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4032 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4033 run a filter in a subprocess. This could lead to confusion in subsequent
4034 lookups in the parent process. There should also be a search_tidyup() at
4035 the end of the subprocess.
4037 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4038 only if the host matched helo_try_verify_hosts, which caused the
4039 verification to occur when the EHLO/HELO command was issued. The ACL just
4040 tested the remembered result. Now, if a previous verification attempt has
4041 not happened, "verify = helo" does it there and then.
4043 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4045 TK/03 Fix log output including CR from clamd.
4047 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4048 exist provoked a memory error which could cause a segfault.
4050 PH/15 Installed PCRE 6.2
4052 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4054 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4055 of the problem. Specifically, suggested +O2 rather than +O1 for the
4058 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4060 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4061 record, the port was not correctly logged when the outgoing_port log
4062 selector was set (it logged the transort's default port).
4064 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4065 fallback_hosts, and "hosts" in the smtp transport.
4067 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4068 host errors such as "Connection refused".
4070 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4071 authentication with radiusclient 0.4.9:
4073 - Error returned from rc_read_config was caught wrongly
4074 - Username/password not passed on to radius server due to wrong length.
4076 The presumption is that some radiusclient API changes for 4.51/PH/17
4077 were not taken care of correctly. The code is still untested by me (my
4078 Linux distribution still has 0.3.2 of radiusclient), but it was
4079 contributed by a Radius user.
4081 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4082 expanding the "port" option of the smtp transport.
4084 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4085 while reading a MIME header. Thanks to Tom Hughes for a patch.
4087 PH/24 Include config.h inside local_scan.h so that configuration settings are
4090 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4091 that in an ACL for RCPT (for example), you can examine exactly what was
4094 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4095 commands, but it was not correctly comparing the address with the actual
4096 client host address. Thus, it would show the EHLO address in Received:
4097 header lines when this was not necessary.
4099 PH/27 Added the % operator to ${eval:}.
4101 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4102 it should be ignoring failures (because with -C, for example, it has lost
4103 privilege). It wasn't ignoring creation failures other than "already
4106 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4109 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4110 because an input line was too long, either on its own, or by virtue of
4111 too many continuations, the temporary file was not being removed, and the
4112 return code was incorrect.
4114 PH/31 Missing "BOOL" in function definition in filtertest.c.
4116 PH/32 Applied Sieve patches from the maintainer.
4118 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4120 PH/33 Added "verify = not_blind".
4122 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4123 Local/Makefile (with some defaults set). These are used in built scripts
4124 such as exicyclog, but they have never been used in the exim_install
4125 script (though there are many overriding facilities there). I have
4126 arranged that the exim_install script now takes note of these two
4129 PH/35 Installed configuration files for Dragonfly.
4131 PH/36 When a locally submitted message by a trusted user did not contain a
4132 From: header, and the sender address was obtained from -f or from an SMTP
4133 MAIL command, and the trusted user did not use -F to supply a sender
4134 name, $originator_name was incorrectly used when constructing a From:
4135 header. Furthermore, $originator_name was used for submission mode
4136 messages from external hosts without From: headers in a similar way,
4137 which is clearly wrong.
4139 PH/37 Added control=suppress_local_fixups.
4141 PH/38 When log_selector = +received_sender was set, and the addition of the
4142 sender made the log line's construction buffer exactly full, or one byte
4143 less than full, an overflow happened when the terminating "\n" was
4146 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4147 when the result of a list match is failure because a DNS lookup failed.
4149 PH/40 RM_COMMAND is now used in the building process.
4151 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4152 the "build-* directories that it finds.
4154 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4155 address in a domain literal was a prefix of an interface address.
4157 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4158 when verifying a sender address, unless rewrite_headers is false.
4160 PH/44 Wrote a long comment about why errors_to addresses are verified as
4161 recipients, not senders.
4163 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4164 the ratelimit ACL was added.
4166 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4168 PH/46 Added extra information about PostgreSQL errors to the error string.
4170 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4171 avoiding the need to install special zones in a real server. This is
4172 backwards compatible; if it can't find the fake resolver, it drops back.
4173 Thus, both old and new test suites can be run.
4175 TF/02 Added util/ratelimit.pl
4177 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4178 clock is set back in time.
4180 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4181 Candler <B.Candler@pobox.com>.
4183 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4184 OK for addresses that are the result of redirections.
4186 PH/48 A number of further additions for the benefit of the new test suite,
4187 including a fake gethostbyname() that interfaces to the fake DNS resolver
4190 TF/06 The fix for widen_domains has also been applied to qualify_single and
4191 search_parents which are the other dnslookup options that can cause
4194 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4197 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4199 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4200 service" if it encountered an MX record with an empty target hostname.
4201 The message is now "an MX or SRV record indicated no SMTP service".
4203 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4204 if the DNS of the sending site is misconfigured. This is quite a
4205 common situation. This change restores the behaviour of treating a
4206 helo verification defer as a failure.
4208 PH/51 If self=fail was set on a router, the bounce message did not include the
4209 actual error message.
4215 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4217 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4218 command itself ended in error, the underlying message about the transport
4219 filter timeout was being overwritten with the pipe command error. Now the
4220 underlying error message should be appended to the second error message.
4222 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4224 PH/02 Reduce the amount of output that "make" produces by default. Full output
4225 can still be requested.
4227 PH/03 The warning log line about a condition test deferring for a "warn" verb
4228 was being output only once per connection, rather than after each
4229 occurrence (because it was using the same function as for successful
4230 "warn" verbs). This seems wrong, so I have changed it.
4232 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4233 it should not have, which might have caused a crash in the right
4234 circumstances, but probably never did.
4236 PH/04 Installed a modified version of Tony Finch's patch to make submission
4237 mode fix the return path as well as the Sender: header line, and to
4238 add a /name= option so that you can make the user's friendly name appear
4241 TF/03 Added the control = fakedefer ACL modifier.
4243 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4244 Mark Lowes for thorough testing.
4246 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4248 TK/03 Merged latest SRS patch from Miles Wilton.
4250 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4251 with the definition in sysexits.h (which is #included earlier).
4252 Fortunately, Exim does not actually use EX_OK. The code used to try to
4253 preserve the sysexits.h value, by assuming that macro definitions were
4254 scanned for macro replacements. I have been disabused of this notion,
4255 so now the code just undefines EX_OK before #including unistd.h.
4257 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4258 in the smtp transport. When a block could not be written in a single
4259 write() function, the timeout was being re-applied to each part-write.
4260 This seems wrong - if the receiver was accepting one byte at a time it
4261 would take for ever. The timeout is now adjusted when this happens. It
4262 doesn't have to be particularly precise.
4264 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4265 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4267 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4268 without a domain if the check to <postmaster@domain> fails.
4270 SC/01 Eximstats: added -xls and the ability to specify output files
4271 (patch written by Frank Heydlauf).
4273 SC/02 Eximstats: use FileHandles for outputting results.
4275 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4277 SC/04 Eximstats: fixed display of large numbers with -nvr option
4279 SC/05 Eximstats: fixed merging of reports with empty tables.
4281 SC/06 Eximstats: added the -include_original_destination flag
4283 SC/07 Eximstats: removed tabs and trailing whitespace.
4285 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4287 TK/06 MBOX spool code: Add real "From " MBOX separator line
4288 so the .eml file is really in mbox format (even though
4289 most programs do not really care). Patch from Alex Miller.
4291 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4292 The latter is generated from $received_to and is only set if the
4293 message has one envelope recipient. SA can use these headers,
4294 obviously out-of-the-box. Patch from Alex Miller.
4296 PH/08 The ${def test on a variable was returning false if the variable's
4297 value was "0", contrary to what the specification has always said!
4298 The result should be true unless the variable is empty.
4300 PH/09 The syntax error of a character other than { following "${if
4301 def:variable_name" (after optional whitespace) was not being diagnosed.
4302 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4303 accidental colon was present, for example, could give incorrect results.
4305 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4306 result is used (not including appendfile, where other changes are about
4309 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4310 This involved changing a lot of size variables from int to off_t. It
4311 should work with maildirs and everything.
4313 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4314 spamd dying while we are connected to it.
4316 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4319 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4320 function for BDB 4.3.
4322 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4324 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4325 the sources for PCRE have been re-arranged and more files are now
4328 PH/15 The code I had for printing potentially long long variables in PH/11
4329 above was not the best (it lost precision). The length of off_t variables
4330 is now inspected at build time, and an appropriate printing format (%ld
4331 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4332 to be "long long int" or "long int". This is needed for the internal
4333 formatting function string_vformat().
4335 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4336 the configuration file to be ":syslog", then the script "guesses" where
4337 the logs files are, rather than using the compiled in default. In our
4338 case the guess is not the same as the compiled default, so the script
4339 suddenly stopped working when I started to use syslog. The patch checks
4340 to see if log_file_path is "". If so, it attempts to read it from exim
4341 with no configuration file to get the compiled in version, before it
4342 falls back to the previous guessing code."
4344 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4345 implementing BATV in an Exim configuration. See NewStuff for the gory
4348 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4349 Makefile that are specific to HP-UX.
4351 PH/18 If the "use_postmaster" option was set for a recipient callout together
4352 with the "random" option, the postmaster address was used as the MAIL
4353 FROM address for the random test, but not for the subsequent recipient
4354 test. It is now used for both.
4356 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4357 patch removes a few documentation additions to RFC 3028, because the
4358 latest draft now contains them. It adds the new en;ascii-case comparator
4359 and a new error check for 8bit text in MIME parts. Comparator and
4360 require names are now matched exactly. I enabled the subaddress
4361 extension, but it is not well tested yet (read: it works for me)."
4363 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4364 rework some of the code of TK/09 above to avoid the hardwired use of
4365 "%lld" and "long long". Replaced the call to snprintf() with a call to
4368 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4369 records point to non-existent hosts", "retry timeout exceeded", and
4370 "retry time not reached for any host after a long failure period".
4372 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4373 experimental DomainKeys support:
4375 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4376 (2) On an error such as an illegally used "control", the wrong name for
4377 the control was given.
4379 These problems did NOT occur unless DomainKeys support was compiled.
4381 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4383 PH/24 Added ${if match_ip condition.
4385 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4386 databases so that it will be absolutely obvious if a crash occurs in the
4387 DB library. This is a regular occurrence (often caused by mis-matched
4390 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4391 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4392 picked up on a user's system that detects such things. There doesn't seem
4393 to be a gcc warning option for this - only an attribute that has to be
4394 put on the function's prototype. It seems that in Fedora Core 4 they have
4395 set this on a number of new functions. No doubt there will be more in due
4398 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4399 specify a transport. However, if an address that was verified by such a
4400 router was the subject of a callout, Exim crashed because it tried to
4401 read the rcpt_include_affixes from the non-existent transport. Now it
4402 just assumes that the setting of that option is false. This bug was
4403 introduced by 4.51/PH/31.
4405 PH/28 Changed -d+all to exclude +memory, because that information is very
4406 rarely of interest, but it makes the output a lot bigger. People tend to
4407 do -d+all out of habit.
4409 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4410 code in os-type was giving problems when libc.so lives in lib64, like on
4413 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4414 aren't the modern standard, and it seems that some systems' include files
4415 don't always have them. Exim was already checking for some of the newer
4416 ones like T_AAAA, and defining it itself. I've added checks for all the
4417 record types that Exim uses.
4419 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4420 not automatically generating a new one, as it is supposed to. This
4421 prevented TLS from working. If the file did exist, but contained invalid
4422 data, a new version was generated, as expected. It was only the case of a
4423 non-existent file that was broken.
4425 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4426 with a change in libdomainkeys > 0.64.
4428 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4429 from DNS. If the selector record carries the flag, it now has
4430 precedence over the domain-wide flag.
4432 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4434 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4435 the use of a port name that isn't defined in /etc/services) occurred, the
4436 message was deferred as in a normal delivery, and thus remained on the
4437 spool, instead of being failed because of the mua_wrapper setting. This
4438 is now fixed, and I tidied up some of the mua_wrapper messages at the
4441 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4442 the messages in a hash of arrays rather than using individual hashes.
4443 This is a bit cleaner and results in dramatic memory savings, albeit
4444 at a slight CPU cost.
4446 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4447 as requested by Marc Sherman.
4449 SC/10 Eximstats: added histograms for user specified patterns as requested
4452 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4454 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4455 fopen() in the content-scanning modules that did not already have it.
4461 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4462 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4464 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4466 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4468 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4469 to test Sieve filters that use "vacation".
4471 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4472 that changes the way the GnuTLS parameters are stored in the cache file.
4473 The new format can be generated externally. For backward compatibility,
4474 if the data in the cache doesn't make sense, Exim assumes it has read an
4475 old-format file, and it generates new data and writes a new file. This
4476 means that you can't go back to an older release without removing the
4479 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4480 work if there are any delivery delays because "one_time" forces the
4481 parent to be marked "delivered", so its unseen clone is never tried
4482 again. For this reason, Exim now forbids the simultaneous setting of
4485 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4486 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4487 entry if you want to know the details. The fix, however, neglected to
4488 consider the case when local delivery batching is involved. The test for
4489 "previously delivered" was not happening when checking to see if an
4490 address could be batched with a previous (undelivered) one; under
4491 certain circumstances this could lead to multiple deliveries to the same
4494 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4495 in its include files, and this causes problems building Exim.
4497 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4498 header_syntax) but Exim was just ignoring anything given after a slash.
4499 In particular, this caused confusion with an attempt to use "verify =
4500 reverse_host_lookup/defer_ok". An error is now given when options are
4501 supplied for verify items that do not have them. (Maybe reverse_host_
4502 lookup should have a defer_ok option, but that's a different point.)
4504 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4505 defined by RFC 821) to 2048, because there were problems with some AUTH
4506 commands, and RFC 1869 says the size should be increased for extended
4507 SMTP commands that take arguments.
4509 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4512 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4513 "unknown" error; now it says that the functionality isn't in the binary.
4515 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4516 an address' error message when a string expansion fails (syntax or
4517 whatever). Otherwise the password may appear in the log. Following change
4518 PH/42 below, there is no longer a chance of it appearing in a bounce
4521 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4523 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4524 of its addresses was ever tried. (Bugzilla bug #2.)
4526 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4527 the result incorrectly in the debug output. (It correctly added a newline
4528 to what was transported.)
4530 TF/01 Added $received_time.
4532 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4533 commented out examples of how to interface to a virus scanner and to
4534 SpamAssassin. Also added commented examples of av_scanner and
4535 spamd_address settings.
4537 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4538 and controls are allowed in which ACLs. There were a couple of minor
4539 errors. Some of the entries in the conditions table (which is a table of
4540 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4541 negation of where the condition IS allowed.
4543 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4545 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4546 header file does not have a version number, so I've had to invent a new
4547 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4548 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4549 radiusclient), but it was contributed by a Radius user.
4551 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4552 files or directories, for OpenSSL.
4554 PH/19 When an Exim process that is running as root has to create an Exim log
4555 file, it does so in a subprocess that runs as exim:exim so as to get the
4556 ownership right at creation (otherwise, other Exim processes might see
4557 the file with the wrong ownership). There was no test for failure of this
4558 fork() call, which would lead to the process getting stuck as it waited
4559 for a non-existent subprocess. Forks do occasionally fail when resources
4560 run out. I reviewed all the other calls to fork(); they all seem to check
4563 PH/20 When checking for unexpected SMTP input at connect time (before writing
4564 the banner), Exim was not dealing correctly with a non-positive return
4565 from the read() function. If the client had disconnected by this time,
4566 the result was a log entry for a synchronization error with an empty
4567 string after "input=" when read() returned zero. If read() returned -1
4568 (an event I could not check), uninitialized data bytes were printed.
4569 There were reports of junk text (parts of files, etc) appearing after
4572 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4574 PH/22 Added support for macro redefinition, and (re)definition in between
4575 driver and ACL definitions.
4577 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4578 forgetting to use the resulting value; it was using the unexpanded value.
4580 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4581 hadn't been configured. The fix is from Juergen Kreileder, who
4582 understands it better than I do:
4584 "Here's what I see happening with three configured cyrus_sasl
4585 authenticators configured (plain, login, cram-md5):
4587 On startup auth_cyrus_sasl_init() gets called for each of these.
4588 This means three calls to sasl_listmech() without a specified mech_list.
4589 => SASL tests which mechs of all available mechs actually work
4590 => three warnings about OTP not working
4591 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4593 With the patch, sasl_listmech() also gets called three times. But now
4594 SASL's mech_list option is set to the server_mech specified in the the
4595 authenticator. Or in other words, the answer from sasl_listmech()
4596 gets limited to just the mech you're testing for (which is different
4598 => the return list contains just 'plain' or 'login', 'cram-md5' or
4599 nothing depending on the value of ob->server_mech.
4601 I've just tested the patch: Authentication still works fine,
4602 unavailable mechs specified in the exim configuration are still
4603 caught, and the auth.log warnings about OTP are gone."
4605 PH/25 When debugging is enabled, the contents of the command line are added
4606 to the debugging output, even when log_selector=+arguments is not
4609 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4610 answer is "GNU", and only if the return is "GNU/something" is the answer
4613 PH/27 $acl_verify_message is now set immediately after the failure of a
4614 verification in an ACL, and so is available in subsequent modifiers. In
4615 particular, the message can be preserved by coding like this:
4617 warn !verify = sender
4618 set acl_m0 = $acl_verify_message
4620 Previously, $acl_verify_message was set only while expanding "message"
4621 and "log_message" when a very denied access.
4623 PH/28 Modified OS/os.c-Linux with
4625 -#ifndef OS_LOAD_AVERAGE
4626 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4628 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4629 nomenclature these days.)
4631 PH/29 Installed patch from the Sieve maintainer that adds the options
4632 sieve_useraddress and sieve_subaddress to the redirect router.
4634 PH/30 In these circumstances:
4635 . Two addresses routed to the same list of hosts;
4636 . First host does not offer TLS;
4637 . First host accepts first address;
4638 . First host gives temporary error to second address;
4639 . Second host offers TLS and a TLS session is established;
4640 . Second host accepts second address.
4641 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4642 and peerdn, if requested) that were in fact used only for the second
4645 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4646 attention to any local part prefix or suffix that was matched by the
4647 router that accepted the address. It now behaves in the same way as it
4648 does for delivery: the affixes are removed from the local part unless
4649 rcpt_include_affixes is set on the transport.
4651 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4652 timeout during the DATA phase of an incoming message.
4654 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4655 applied a patch sanctioned by the Sieve maintainer.
4657 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4658 the uid or gid is negative. A case of a negative gid caused this to be
4659 noticed. The fix allows for either to be negative.
4661 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4662 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4665 PH/36 The change PH/12 above was broken. Fixed it.
4667 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4668 the grounds that routing the same address twice would always produce the
4669 same answer. This might have been true once, but it is certainly no
4670 longer true now. Routing a child address may depend on the previous
4671 routing that produced that child. Some complicated redirection strategies
4672 went wrong when messages had multiple recipients, and made Exim's
4673 behaviour dependent on the order in which the addresses were given.
4675 I have moved the duplicate checking until after the routing is complete.
4676 Exim scans the addresses that are assigned to local and remote
4677 transports, and removes any duplicates. This means that more work will be
4678 done, as duplicates will always all be routed, but duplicates are
4679 presumably rare, so I don't expect this is of any significance.
4681 For deliveries to pipes, files, and autoreplies, the duplicate checking
4682 still happens during the routing process, since they are not going to be
4685 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4686 It corrects a timeout issue with spamd. This is Ian's comment: "The
4687 background is that sometimes spamd either never reads data from a
4688 connection it has accepted, or it never writes response data. The exiscan
4689 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4690 blindly assumes that writes won't block so it may never time out."
4692 PH/39 Allow G after quota size as well as K and M.
4694 PH/40 The value set for $authenticated_id in an authenticator may not contain
4695 binary zeroes or newlines because the value is written to log lines and
4696 to spool files. There was no check on this. Now the value is run through
4697 the string_printing() function so that such characters are converted to
4698 printable escape sequences.
4700 PH/41 $message_linecount is a new variable that contains the total number of
4701 lines in the message. Compare $body_linecount, which is the count for the
4704 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4705 bounce and delay warning messages, except in certain special cases, which
4708 (a) An SMTP error message from a remote host;
4709 (b) A message specified in a :fail: redirection;
4710 (c) A message specified in a "fail" command in a system filter;
4711 (d) A message specified in a FAIL return from the queryprogram router;
4712 (e) A message specified by the cannot_route_message router option.
4714 In these cases only, Exim does include the error details in bounce and
4715 warning messages. There are also a few cases where bland messages such
4716 as "unrouteable address" or "local delivery error" are given.
4718 PH/43 $value is now also set for the "else" part of a ${run expansion.
4720 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4721 being worked on, but at least Exim now implements the latest version to
4724 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4725 process to complete was treated as a delivery failure, a timeout while
4726 writing the message to the pipe was logged, but erroneously treated as a
4727 successful delivery. Such timeouts include transport filter timeouts. For
4728 consistency with the overall process timeout, these timeouts are now
4729 treated as errors, giving rise to delivery failures by default. However,
4730 there is now a new Boolean option for the pipe transport called
4731 timeout_defer, which, if set TRUE, converts the failures into defers for
4732 both kinds of timeout. A transport filter timeout is now identified in
4735 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4736 systems where "make" and "gmake" are different, calling "gmake" at top
4737 level broke things. I've arranged for the value of $(MAKE) to be passed
4738 from the Makefile to this script so that it can call the same version of
4742 A note about Exim versions 4.44 and 4.50
4743 ----------------------------------------
4745 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4746 changes of various kinds. As a consequence, a big documentation update was
4747 needed. This delayed the release for rather longer than seemed good, especially
4748 in the light of a couple of (minor) security issues. Therefore, the changes
4749 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4750 release. So 4.44 and 4.50 are in effect two different branches that both start
4753 I have left the 4.50 change log unchanged; it contains all the changes since
4754 4.43. The change log for 4.44 is below; many of its items are identical to
4755 those for 4.50. This seems to be the most sensible way to preserve the
4756 historical information.
4762 1. Minor wording change to the doc/README.SIEVE file.
4764 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4765 computation of the current number of files was incorrect.
4767 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4768 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4771 4. Give more explanation in the error message when the command for a transport
4772 filter fails to execute.
4774 5. There are several places where Exim runs a non-Exim command in a
4775 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4776 was being done only for the command run by the queryprogram router. It is
4777 now done for all such subprocesses. The other cases are: ${run, transport
4778 filters, and the commands run by the lmtp and pipe transports.
4780 6. Added CONFIGURE_GROUP build-time option.
4782 7. Some older OS have a limit of 256 on the maximum number of file
4783 descriptors. Exim was using setrlimit() to set 1000 as a large value
4784 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4785 systems. I've change it so that if it can't get 1000, it tries for 256.
4787 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4788 was an oversight, and furthermore, ever since the addition of extra
4789 controls (e.g. 4.43/32), the checks on when to allow different forms of
4790 "control" were broken. There should now be diagnostics for all cases when a
4791 control that does not make sense is encountered.
4793 9. Added the /retain_sender option to "control=submission".
4795 10. $recipients is now available in the predata ACL (oversight).
4797 11. Tidy the search cache before the fork to do a delivery from a message
4798 received from the command line. Otherwise the child will trigger a lookup
4799 failure and thereby defer the delivery if it tries to use (for example) a
4800 cached ldap connection that the parent has called unbind on.
4802 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4803 of $address_data from the recipient verification was clobbered by the
4804 sender verification.
4806 13. The value of address_data from a sender verification is now available in
4807 $sender_address_data in subsequent conditions in the ACL statement.
4809 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4811 15. Added a new option "connect=<time>" to callout options, to set a different
4814 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4815 was its contents. (It was OK if the option was not defined at all.)
4817 17. A "Completed" log line is now written for messages that are removed from
4818 the spool by the -Mrm option.
4820 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4821 information about exactly what failed.
4823 19. Added -dd to debug only the daemon process.
4825 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4826 handles timeouts, both on the server side and network timeouts. Renamed the
4827 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4829 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4830 It is now set to "smtps".
4832 22. $host_address is now set to the target address during the checking of
4833 ignore_target_hosts.
4835 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4836 being passed; this would have caused $sender_host_name to have been used if
4837 matching the list had actually called for a host name (not very likely,
4838 since this list is usually IP addresses). A host name is now passed as
4841 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4842 code that specifies a non-restarting handler (typically sigaction() in
4843 modern systems) in an attempt to fix a rare and obscure crash bug.
4845 25. Narrowed the window for a race in the daemon that could cause it to ignore
4846 SIGCHLD signals. This is not a major problem, because they are used only to
4847 wake it up if nothing else does.
4849 26. A malformed maildirsize file could cause Exim to calculate negative values
4850 for the mailbox size or file count. Odd effects could occur as a result.
4851 The maildirsize information is now recalculated if the size or filecount
4854 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4855 support for a long time. Removed HAVE_SYS_VFS_H.
4857 28. Installed the latest version of exipick from John Jetmore.
4859 29. In an address list, if the pattern was not a regular expression, an empty
4860 subject address (from a bounce message) matched only if the pattern was an
4861 empty string. Non-empty patterns were not even tested. This was the wrong
4862 because it is perfectly reasonable to use an empty address as part of a
4863 database query. An empty address is now tested by patterns that are
4864 lookups. However, all the other forms of pattern expect the subject to
4865 contain a local part and a domain, and therefore, for them, an empty
4866 address still always fails if the pattern is not itself empty.
4868 30. Exim went into a mad DNS loop when attempting to do a callout where the
4869 host was specified on an smtp transport, and looking it up yielded more
4870 than one IP address.
4872 31. Re-factored the code for checking spool and log partition space into a
4873 function that finds that data and another that does the check. The former
4874 is then used to implement four new variables: $spool_space, $log_space,
4875 $spool_inodes, and $log_inodes.
4877 32. The RFC2047 encoding function was originally intended for short strings
4878 such as real names; it was not keeping to the 75-character limit for
4879 encoded words that the RFC imposes. It now respects the limit, and
4880 generates multiple encoded words if necessary. To be on the safe side, I
4881 have increased the buffer size for the ${rfc2047: expansion operator from
4884 33. It is now permitted to omit both strings after an "if" condition; if the
4885 condition is true, the result is "true". As before, when the second string
4886 is omitted, a false condition yields an empty string. This makes it less
4887 cumbersome to write custom ACL and router conditions.
4889 34. Failure to deliver a bounce message always caused it to be frozen, even if
4890 there was an errors_to setting on the router. The errors_to setting is now
4893 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4894 canonical form (fully expanded) before being placed in
4895 $sender_host_address.
4897 36. The table in the code that translates DNS record types into text (T_A to
4898 "A" for instance) was missing entries for NS and CNAME. It is just possible
4899 that this could have caused confusion if both these types were looked up
4900 for the same domain, because the text type is used as part of Exim's
4901 per-process caching. But the chance of anyone hitting this buglet seems
4904 37. The dnsdb lookup has been extended in a number of ways.
4906 (1) There is a new type, "zns", which walks up the domain tree until it
4907 finds some nameserver records. It should be used with care.
4909 (2) There is a new type, "mxh", which is like "mx" except that it returns
4910 just the host names, not the priorities.
4912 (3) It is now possible to give a list of domains (or IP addresses) to be
4913 looked up. The behaviour when one of the lookups defers can be
4914 controlled by a keyword.
4916 (4) It is now possible to specify the separator character for use when
4917 multiple records are returned.
4919 38. The dnslists ACL condition has been extended: it is now possible to supply
4920 a list of IP addresses and/or domains to be looked up in a particular DNS
4923 39. Added log_selector=+queue_time_overall.
4925 40. When running the queue in the test harness, wait just a tad after forking a
4926 delivery process, to get repeatability of debugging output.
4928 41. Include certificate and key file names in error message when GnuTLS fails
4929 to set them up, because the GnuTLS error message doesn't include the name
4930 of the failing file when there is a problem reading it.
4932 42. Allow both -bf and -bF in the same test run.
4934 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4936 44. The "Exiscan patch" is now merged into the mainline Exim source.
4938 45. Sometimes the final signoff response after QUIT could fail to get
4939 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4940 < 0 before doing a fflush(). This bug looks as though it goes back to the
4941 introduction of TLS in release 3.20, but "sometimes" must have been rare
4942 because the tests only now provoked it.
4944 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4945 (this can affect the format of dates).
4947 47. exim_tidydb, when checking for the continued existence of a message for
4948 which it has found a message-specific retry record, was not finding
4949 messages that were in split spool directories. Consequently, it was
4950 deleting retry records that should have stayed in existence.
4952 48. Steve fixed some bugs in eximstats.
4954 49. The SPA authentication driver was not abandoning authentication and moving
4955 on to the next authenticator when an expansion was forced to fail,
4956 contradicting the general specification for all authenticators. Instead it
4957 was generating a temporary error. It now behaves as specified.
4959 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4960 (the order specifies the preference for clients). The order is now AES256,
4961 AES128, 3DES, ARCFOUR128.
4963 51. Small patch to Sieve code - explicitly set From: when generating an
4966 52. Exim crashed if a remote delivery caused a very long error message to be
4967 recorded - for instance if somebody sent an entire SpamAssassin report back
4968 as a large number of 550 error lines. This bug was coincidentally fixed by
4969 increasing the size of one of Exim's internal buffers (big_buffer) that
4970 happened as part of the Exiscan merge. However, to be on the safe side, I
4971 have made the code more robust (and fixed the comments that describe what
4974 53. Now that there can be additional text after "Completed" in log lines (if
4975 the queue_time_overall log selector is set), a one-byte patch to exigrep
4976 was needed to allow it to recognize "Completed" as not the last thing in
4979 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4980 patch that reportedly fixes this has been added. I am not expert enough to
4981 create a test for it. This is what the patch creator wrote:
4983 "I found a little strange behaviour of ldap code when working with
4984 Windows 2003 AD Domain, where users was placed in more than one
4985 Organization Units. When I tried to give exim partial DN, the exit code
4986 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4987 But simultaneously result of request was absolutely normal ldap result,
4988 so I produce this patch..."
4990 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4991 REFERENCE, so I have modified the code to exclude the patch when that macro
4994 55. Some experimental protocols are using DNS PTR records for new purposes. The
4995 keys for these records are domain names, not reversed IP addresses. The
4996 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4997 leaves it alone. Component reversal etc. now happens only for IP addresses.
5000 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5002 57. Double the size of the debug message buffer (to 2048) so that more of very
5003 long debug lines gets shown.
5005 58. The exicyclog utility now does better if the number of log files to keep
5006 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5008 59. Two changes related to the smtp_active_hostname option:
5010 (1) $smtp_active_hostname is now available as a variable.
5011 (2) The default for smtp_banner uses $smtp_active_hostname instead
5012 of $primary_hostname.
5014 60. The host_aton() function is supposed to be passed a string that is known
5015 to be a valid IP address. However, in the case of IPv6 addresses, it was
5016 not checking this. This is a hostage to fortune. Exim now panics and dies
5017 if the condition is not met. A case was found where this could be provoked
5018 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5019 components; fortuitously, this particular loophole had already been fixed
5020 by change 4.50/55 above.
5022 If there are any other similar loopholes, the new check in host_aton()
5023 itself should stop them being exploited. The report I received stated that
5024 data on the command line could provoke the exploit when Exim was running as
5025 exim, but did not say which command line option was involved. All I could
5026 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5027 running as the user.
5030 61. There was a buffer overflow vulnerability in the SPA authentication code
5031 (which came originally from the Samba project). I have added a test to the
5032 spa_base64_to_bits() function which I hope fixes it.
5035 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5036 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5038 63. The daemon start-up calls getloadavg() while still root for those OS that
5039 need the first call to be done as root, but it missed one case: when
5040 deliver_queue_load_max is set with deliver_drop_privilege. This is
5041 necessary for the benefit of the queue runner, because there is no re-exec
5042 when deliver_drop_privilege is set.
5044 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5045 This has been fixed.
5047 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5048 was in use, was not putting the data itself into the right store pool;
5049 consequently, it could be overwritten for a subsequent message in the same
5050 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5053 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5055 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5056 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5057 the function were treating the return as a boolean value, which happened to
5058 work because 0=false and not-0=true, but is not correct code.
5060 68. The host_aton() function was not handling scoped IPv6 addresses (those
5061 with, for example, "%eth0" on the end) correctly.
5063 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5064 negated items (that is, ~something) in unsigned ints. Some compilers
5065 apparently mutter when there is no cast.
5067 70. If an address verification called from an ACL failed, and did not produce a
5068 user-specific message (i.e. there was only a "system" message), nothing was
5069 put in $acl_verify_message. In this situation, it now puts the system
5072 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5073 session; change 4.31/43 added the unwanted input to the log line - except
5074 that it did not do this in the start of session case. It now does.
5076 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5077 This is wrong and can cause the other end to generate a synchronization
5078 error if it is another Exim or anything else that does the synchronization
5079 check. A QUIT command is no longer sent after a timeout.
5081 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5082 during host lookups.
5084 74. The defer_ok option of callout verification was not working if it was used
5085 when verifying addresses in header lines, that is, for this case:
5087 verify = header_sender/callout=defer_ok
5089 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5090 those file descriptors could be used for SMTP connections. If anything
5091 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5092 could be sent to the SMTP client, causing chaos. The daemon now opens
5093 stdin, stdout, and stderr to /dev/null when it puts itself into the
5096 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5097 log by default. The user can override this with suitable Perl magic.
5099 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5100 the log message when discard triggers, was not working for the DATA ACL or
5101 for the non-SMTP ACL.
5103 78. Error message wording change in sieve.c.
5105 79. If smtp_accept_max_per_host was set, the number of connections could be
5106 restricted to fewer than expected, because the daemon was trying to set up
5107 a new connection before checking whether the processes handling previous
5108 connections had finished. The check for completed processes is now done
5109 earlier. On busy systems, this bug wouldn't be noticed because something
5110 else would have woken the daemon, and it would have reaped the completed
5113 80. If a message was submitted locally by a user whose login name contained one
5114 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5115 It caused a spool format error. I have fixed the spool reading code. A
5116 related problem was that the "from" clause in the Received: line became
5117 illegal because of the space(s). It is now covered by ${quote_local_part.
5119 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5122 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5124 83. Patch from Sieve maintainer to fix unterminated string problem in
5125 "vacation" handling.
5127 84. Some minor changes to the Linux configuration files to help with other
5128 OS variants using glibc.
5130 85. One more patch for Sieve to update vacation handling to latest spec.
5133 ----------------------------------------------------
5134 See the note above about the 4.44 and 4.50 releases.
5135 ----------------------------------------------------
5141 1. Change 4.43/35 introduced a bug that caused file counts to be
5142 incorrectly computed when quota_filecount was set in an appendfile
5145 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5146 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5149 3. Give more explanation in the error message when the command for a transport
5150 filter fails to execute.
5152 4. There are several places where Exim runs a non-Exim command in a
5153 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5154 was being done only for the command run by the queryprogram router. It is
5155 now done for all such subprocesses. The other cases are: ${run, transport
5156 filters, and the commands run by the lmtp and pipe transports.
5158 5. Some older OS have a limit of 256 on the maximum number of file
5159 descriptors. Exim was using setrlimit() to set 1000 as a large value
5160 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5161 systems. I've change it so that if it can't get 1000, it tries for 256.
5163 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5164 was an oversight, and furthermore, ever since the addition of extra
5165 controls (e.g. 4.43/32), the checks on when to allow different forms of
5166 "control" were broken. There should now be diagnostics for all cases when a
5167 control that does not make sense is encountered.
5169 7. $recipients is now available in the predata ACL (oversight).
5171 8. Tidy the search cache before the fork to do a delivery from a message
5172 received from the command line. Otherwise the child will trigger a lookup
5173 failure and thereby defer the delivery if it tries to use (for example) a
5174 cached ldap connection that the parent has called unbind on.
5176 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5177 of $address_data from the recipient verification was clobbered by the
5178 sender verification.
5180 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5181 was its contents. (It was OK if the option was not defined at all.)
5183 11. A "Completed" log line is now written for messages that are removed from
5184 the spool by the -Mrm option.
5186 12. $host_address is now set to the target address during the checking of
5187 ignore_target_hosts.
5189 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5190 being passed; this would have caused $sender_host_name to have been used if
5191 matching the list had actually called for a host name (not very likely,
5192 since this list is usually IP addresses). A host name is now passed as
5195 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5196 code that specifies a non-restarting handler (typically sigaction() in
5197 modern systems) in an attempt to fix a rare and obscure crash bug.
5199 15. Narrowed the window for a race in the daemon that could cause it to ignore
5200 SIGCHLD signals. This is not a major problem, because they are used only to
5201 wake it up if nothing else does.
5203 16. A malformed maildirsize file could cause Exim to calculate negative values
5204 for the mailbox size or file count. Odd effects could occur as a result.
5205 The maildirsize information is now recalculated if the size or filecount
5208 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5209 support for a long time. Removed HAVE_SYS_VFS_H.
5211 18. Updated exipick to current release from John Jetmore.
5213 19. Allow an empty sender to be matched against a lookup in an address list.
5214 Previously the only cases considered were a regular expression, or an
5217 20. Exim went into a mad DNS lookup loop when doing a callout where the
5218 host was specified on the transport, if the DNS lookup yielded more than
5221 21. The RFC2047 encoding function was originally intended for short strings
5222 such as real names; it was not keeping to the 75-character limit for
5223 encoded words that the RFC imposes. It now respects the limit, and
5224 generates multiple encoded words if necessary. To be on the safe side, I
5225 have increased the buffer size for the ${rfc2047: expansion operator from
5228 22. Failure to deliver a bounce message always caused it to be frozen, even if
5229 there was an errors_to setting on the router. The errors_to setting is now
5232 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5233 canonical form (fully expanded) before being placed in
5234 $sender_host_address.
5236 24. Updated eximstats to version 1.33
5238 25. Include certificate and key file names in error message when GnuTLS fails
5239 to set them up, because the GnuTLS error message doesn't include the name
5240 of the failing file when there is a problem reading it.
5242 26. Expand error message when OpenSSL has problems setting up cert/key files.
5245 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5246 (this can affect the format of dates).
5248 28. exim_tidydb, when checking for the continued existence of a message for
5249 which it has found a message-specific retry record, was not finding
5250 messages that were in split spool directories. Consequently, it was
5251 deleting retry records that should have stayed in existence.
5253 29. eximstats updated to version 1.35
5254 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5255 1.35 - bugfix such that pie charts by volume are generated correctly
5257 30. The SPA authentication driver was not abandoning authentication and moving
5258 on to the next authenticator when an expansion was forced to fail,
5259 contradicting the general specification for all authenticators. Instead it
5260 was generating a temporary error. It now behaves as specified.
5262 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5263 (the order specifies the preference for clients). The order is now AES256,
5264 AES128, 3DES, ARCFOUR128.
5266 31. Small patch to Sieve code - explicitly set From: when generating an
5269 32. Exim crashed if a remote delivery caused a very long error message to be
5270 recorded - for instance if somebody sent an entire SpamAssassin report back
5271 as a large number of 550 error lines. This bug was coincidentally fixed by
5272 increasing the size of one of Exim's internal buffers (big_buffer) that
5273 happened as part of the Exiscan merge. However, to be on the safe side, I
5274 have made the code more robust (and fixed the comments that describe what
5277 33. Some experimental protocols are using DNS PTR records for new purposes. The
5278 keys for these records are domain names, not reversed IP addresses. The
5279 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5280 leaves it alone. Component reversal etc. now happens only for IP addresses.
5283 34. The host_aton() function is supposed to be passed a string that is known
5284 to be a valid IP address. However, in the case of IPv6 addresses, it was
5285 not checking this. This is a hostage to fortune. Exim now panics and dies
5286 if the condition is not met. A case was found where this could be provoked
5287 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5288 components; fortuitously, this particular loophole had already been fixed
5289 by change 4.50/55 or 4.44/33 above.
5291 If there are any other similar loopholes, the new check in host_aton()
5292 itself should stop them being exploited. The report I received stated that
5293 data on the command line could provoke the exploit when Exim was running as
5294 exim, but did not say which command line option was involved. All I could
5295 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5296 running as the user.
5299 35. There was a buffer overflow vulnerability in the SPA authentication code
5300 (which came originally from the Samba project). I have added a test to the
5301 spa_base64_to_bits() function which I hope fixes it.
5304 36. The daemon start-up calls getloadavg() while still root for those OS that
5305 need the first call to be done as root, but it missed one case: when
5306 deliver_queue_load_max is set with deliver_drop_privilege. This is
5307 necessary for the benefit of the queue runner, because there is no re-exec
5308 when deliver_drop_privilege is set.
5310 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5311 was in use, was not putting the data itself into the right store pool;
5312 consequently, it could be overwritten for a subsequent message in the same
5313 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5316 38. Sometimes the final signoff response after QUIT could fail to get
5317 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5318 < 0 before doing a fflush(). This bug looks as though it goes back to the
5319 introduction of TLS in release 3.20, but "sometimes" must have been rare
5320 because the tests only now provoked it.
5326 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5327 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5328 Now it returns an integer. A number of calls were still expecting a T/F
5329 return. Fortuitously, in all cases, the tests worked in OK situations,
5330 which is the norm. However, things would have gone wrong on any write
5331 failures on the smtp file descriptor. This function is used when sending
5332 messages over SMTP and also when doing verify callouts.
5334 2. When Exim is called to do synchronous delivery of a locally submitted
5335 message (the -odf or -odi options), it no longer closes stderr before doing
5338 3. Implemented the mua_wrapper option.
5340 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5342 5. Implemented the functions header_remove(), header_testname(),
5343 header_add_at_position(), and receive_remove_recipient(), and exported them
5346 6. If an ACL "warn" statement specified the addition of headers, Exim already
5347 inserted X-ACL-Warn: at the start if there was no header name. However, it
5348 was not making this test for the second and subsequent header lines if
5349 there were newlines in the string. This meant that an invalid header could
5350 be inserted if Exim was badly configured.
5352 7. Allow an ACL "warn" statement to add header lines at the start or after all
5353 the Received: headers, as well as at the end.
5355 8. Added the rcpt_4xx retry error code.
5357 9. Added postmaster_mailfrom=xxx to callout verification option.
5359 10. Added mailfrom=xxxx to the callout verification option, for verify=
5362 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5363 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5365 12. Inserted some casts to stop certain compilers warning when using pointer
5366 differences as field lengths or precisions in printf-type calls (mostly
5367 affecting debugging statements).
5369 13. Added optional readline() support for -be (dynamically loaded).
5371 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5372 same clock tick as a message's arrival, so that its received time was the
5373 same as the "first fail" time on the retry record, and that message
5374 remained on the queue past the ultimate address timeout, every queue runner
5375 would try a delivery (because it was past the ultimate address timeout) but
5376 after another failure, the ultimate address timeout, which should have then
5377 bounced the address, did not kick in. This was a "< instead of <=" error;
5378 in most cases the first failure would have been in the next clock tick
5379 after the received time, and all would be well.
5381 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5382 being recognized when the domain list was tested by the match_domain
5383 condition in an expansion string.
5385 16. Added the ${str2b64: operator.
5387 17. Exim was always calling setrlimit() to set a large limit for the number of
5388 processes, without checking whether the existing limit was already
5389 adequate. (It did check for the limit on file descriptors.) Furthermore,
5390 errors from getrlimit() and setrlimit() were being ignored. Now they are
5391 logged to the main and panic logs, but Exim does carry on, to try to do its
5392 job under whatever limits there are.
5394 18. Imported PCRE 5.0.
5396 19. Trivial typo in log message " temporarily refused connection" (the leading
5399 20. If the log selector return_path_on_delivery was set and an address was
5400 redirected to /dev/null, the delivery process crashed because it assumed
5401 that a return path would always be set for a "successful" delivery. In this
5402 case, the whole delivery is bypassed as an optimization, and therefore no
5405 21. Internal re-arrangement: the function for sending a challenge and reading
5406 a response while authentication was assuming a zero-terminated challenge
5407 string. It's now changed to take a pointer and a length, to allow for
5408 binary data in such strings.
5410 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5412 23. Exim was not respecting finduser_retries when seeking the login of the
5413 uid under which it was called; it was always trying 10 times. (The default
5414 setting of finduser_retries is zero.) Also, it was sleeping after the final
5415 failure, which is pointless.
5417 24. Implemented tls_on_connect_ports.
5419 25. Implemented acl_smtp_predata.
5421 26. If the domain in control=submission is set empty, Exim assumes that the
5422 authenticated id is a complete email address when it generates From: or
5423 Sender: header lines.
5425 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5426 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5427 chown and chgrp in /bin and hostname in /usr/bin.
5429 28. Exim was keeping the "process log" file open after each use, just as it
5430 does for the main log. This opens the possibility of it remaining open for
5431 long periods when the USR1 signal hits a daemon. Occasional processlog
5432 errors were reported, that could have been caused by this. Anyway, it seems
5433 much more sensible not to leave this file open at all, so that is what now
5436 29. The long-running daemon process does not normally write to the log once it
5437 has entered its main loop, and it closes the log before doing so. This is
5438 so that log files can straightforwardly be renamed and moved. However,
5439 there are a couple of unusual error situations where the daemon does write
5440 log entries, and I had neglected to close the log afterwards.
5442 30. The text of an SMTP error response that was received during a remote
5443 delivery was being truncated at 512 bytes. This is too short for some of
5444 the long messages that one sometimes sees. I've increased the limit to
5447 31. It is now possible to make retry rules that apply only when a message has a
5448 specific sender, in particular, an empty sender.
5450 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5451 it possible to be selective about when SMTP synchronization is enforced.
5453 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5455 32. Implemented hosts_connection_nolog.
5457 33. Added an ACL for QUIT.
5459 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5462 35. Added mailbox_size and mailbox_filecount to appendfile.
5464 36. Added control = no_multiline_responses to ACLs.
5466 37. There was a bug in the logic of the code that waits for the clock to tick
5467 in the case where the clock went backwards by a substantial amount such
5468 that the microsecond fraction of "now" was more than the microsecond
5469 fraction of "then" (but the whole seconds number was less).
5471 38. Added support for the libradius Radius client library this is found on
5472 FreeBSD (previously only the radiusclient library was supported).
5478 1. When certain lookups returned multiple values in the form name=value, the
5479 quoting of the values was not always being done properly. Specifically:
5480 (a) If the value started with a double quote, but contained no whitespace,
5482 (b) If the value contained whitespace other than a space character (i.e.
5483 tabs or newlines or carriage returns) it was not quoted.
5484 This fix has been applied to the mysql and pgsql lookups by writing a
5485 separate quoting function and calling it from the lookup code. The fix
5486 should probably also be applied to nisplus, ibase and oracle lookups, but
5487 since I cannot test any of those, I have not disturbed their existing code.
5489 2. A hit in the callout cache for a specific address caused a log line with no
5490 reason for rejecting RCPT. Now it says "Previous (cached) callout
5491 verification failure".
5493 3. There was an off-by-one bug in the queryprogram router. An over-long
5494 return line was truncated at 256 instead of 255 characters, thereby
5495 overflowing its buffer with the terminating zero. As well as fixing this, I
5496 have increased the buffer size to 1024 (and made a note to document this).
5498 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5499 when Exim is waiting for an SMTP response from a remote server, Exim
5500 restarts its select() call on the socket, thereby resetting its timeout.
5501 This is not a problem when such interrupts are rare. Somebody set up a cron
5502 job to run exiwhat every 2 minutes, which is less than the normal select()
5503 timeout (5 or 10 minutes). This meant that the select() timeout never
5504 kicked in because it was always reset. I have fixed this by comparing the
5505 time when an interrupt arrives with the time at the start of the first call
5506 to select(). If more time than the timeout has elapsed, the interrupt is
5507 treated as a timeout.
5509 5. Some internal re-factoring in preparation for the addition of Sieve
5510 extensions (by MH). In particular, the "personal" test is moved to a
5511 separate function, and given an option for scanning Cc: and Bcc: (which is
5512 not set for Exim filters).
5514 6. When Exim created an email address using the login of the caller as the
5515 local part (e.g. when creating a From: or Sender: header line), it was not
5516 quoting the local part when it contained special characters such as @.
5518 7. Installed new OpenBSD configuration files.
5520 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5521 try to make them clearer.
5523 9. Callout options, other than the timeout value, were being ignored when
5524 verifying sender addresses in header lines. For example, when using
5526 verify = header_sender/callout=no_cache
5528 the cache was (incorrectly) being used.
5530 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5531 only the Cygwin environment.
5533 11. When return_path_on_delivery was set as a log selector, if different remote
5534 addresses in the same message used different return paths and parallel
5535 remote delivery occurred, the wrong values would sometimes be logged.
5536 (Whenever a remote delivery process finished, the return path value from
5537 the most recently started remote delivery process was logged.)
5539 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5540 lines when AUTH and/or TLS are in use. This is the "received protocol"
5541 field. Exim used to use "asmtp" for authenticated SMTP, without any
5542 indication (in the protocol name) for TLS use. Now it follows the RFC and
5543 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5544 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5545 names appear in log lines as well as in Received: header lines.
5547 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5548 extensions, and comparison tests, and to fix some bugs.
5550 14. Changes to the "personal" filter test:
5552 (1) The test was buggy in that it was just doing the equivalent of
5553 "contains" tests on header lines. For example, if a user's address was
5554 anne@some.where, the "personal" test would incorrectly be true for
5556 To: susanne@some.where
5558 This test is now done by extracting each address from the header in turn,
5559 and checking the entire address. Other tests that are part of "personal"
5560 are now done using regular expressions (for example, to check local parts
5561 of addresses in From: header lines).
5563 (2) The list of non-personal local parts in From: addresses has been
5564 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5565 taken from the Sieve specification recommendations.
5567 (3) If the message contains any header line starting with "List-" it is
5568 treated as non-personal.
5570 (4) The test for "circular" in the Subject: header line has been removed
5571 because it now seems ill-conceived.
5573 15. Minor typos in src/EDITME comments corrected.
5575 16. Installed latest exipick from John Jetmore.
5577 17. If headers_add on a router specified a text string that was too long for
5578 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5579 of string_sprintf() is now avoided.
5581 18. $message_body_size was not set (it was always zero) when running the DATA
5582 ACL and the local_scan() function.
5584 19. For the "mail" command in an Exim filter, no default was being set for
5585 the once_repeat time, causing a random time value to be used if "once" was
5586 specified. (If the value happened to be <= 0, no repeat happened.) The
5587 default is now 0s, meaning "never repeat". The "vacation" command was OK
5588 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5589 (I found it when inspecting the code).
5591 20. There is now an overall timeout for performing a callout verification. It
5592 defaults to 4 times the callout timeout, which applies to individual SMTP
5593 commands during the callout. The overall timeout applies when there is more
5594 than one host that can be tried. The timeout is checked before trying the
5595 next host. This prevents very long delays if there are a large number of
5596 hosts and all are timing out (e.g. when the network connections are timing
5597 out). The value of the overall timeout can be changed by specifying an
5598 additional sub-option for "callout", called "maxwait". For example:
5600 verify = sender/callout=5s,maxwait=20s
5602 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5603 to the end before writing, but this should make it even safer).
5605 22. Exim was forgetting that it had advertised PIPELINING for the second and
5606 subsequent messages on an SMTP connection. It was also not resetting its
5607 memory on STARTTLS and an internal HELO.
5609 23. When Exim logs an SMTP synchronization error within a session, it now
5610 records whether PIPELINING has been advertised or not.
5612 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5613 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5614 rather than long int.
5616 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5618 26. Added the never_mail option to autoreply.
5624 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5625 crash if the getsockname() call failed; this can happen if a connection is
5626 closed very soon after it is established. The problem was simply in the
5627 order in which certain operations were done, causing Exim to try to write
5628 to the SMTP stream before it had set up the file descriptor. The bug has
5629 been fixed by making things happen in the correct order.
5635 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5636 before the connection was closed, thus losing the rejection response.
5638 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5639 some early Solaris releases, but causes trouble in current releases where
5640 socklen_t is defined.
5642 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5645 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5648 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5649 that releasing the top of it at the end releases what was used for sub-
5650 expansions (unless the block got too big). However, discard this block if
5651 the first thing is a variable or header, so that we can use its block when
5652 it is dynamic (useful for very large $message_headers, for example).
5654 6. Lookups now cache *every* query, not just the most recent. A new, separate
5655 store pool is used for this. It can be recovered when all lookup caches are
5656 flushed. Lookups now release memory at the end of their result strings.
5657 This has involved some general refactoring of the lookup sources.
5659 7. Some code has been added to the store_xxx() functions to reduce the amount
5660 of flapping under certain conditions.
5662 8. log_incoming_interface used to affect only the <= reception log lines. Now
5663 it causes the local interface and port to be added to several more SMTP log
5664 lines, for example "SMTP connection from", and rejection lines.
5666 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5668 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5670 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5671 could be overwritten at the end of the current message (or the start of a
5672 new message if it was set in a HELO ACL). The value is now preserved for
5673 the duration of the SMTP connection.
5675 12. If a transport had a headers_rewrite setting, and a matching header line
5676 contained an unqualified address, that address was qualified, even if it
5677 did not match any rewriting rules. The underlying bug was that the values
5678 of the flags that permit the existence of unqualified sender and recipient
5679 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5680 non-local messages, and by -bnq for local messages) were not being
5681 preserved with the message after it was received.
5683 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5684 "next input=" as part of the text comprising the host identity instead of
5685 the correct text. The code was using the same buffer for two different
5686 strings. However, depending on which order the printing function evaluated
5687 its arguments, the bug did not always show up. Under Linux, for example, my
5688 test suite worked just fine.
5690 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5691 doesn't work with some older versions of Perl. It has been changed to "my",
5692 which in any case is probably the better facility to use.
5694 15. A really picky compiler found some instances of statements for creating
5695 error messages that either had too many or two few arguments for the format
5698 16. The size of the buffer for calls to the DNS resolver has been increased
5699 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5700 for addresses that have a lot of PTR records. This alleviates a problem; it
5701 does not fully solve it.
5703 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5704 buffer now truncates the list and logs the incident, which is the same
5705 action as happens when Exim is looking up a host name and its aliases.
5706 Previously in this situation something unpredictable would happen;
5707 sometimes it was "internal error: store_reset failed".
5709 18. If a server dropped the connection unexpectedly when an Exim client was
5710 using GnuTLS and trying to read a response, the client delivery process
5711 crashed while trying to generate an error log message.
5713 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5714 string, for example:
5716 warn message = H1: something\nH2: something
5718 the text was added as a single header line from Exim's point of view
5719 though it ended up OK in the delivered message. However, searching for the
5720 second and subsequent header lines using $h_h2: did not work. This has been
5721 fixed. Similarly, if a system filter added multiple headers in this way,
5722 the routers could not see them.
5724 20. Expanded the error message when iplsearch is called with an invalid key to
5725 suggest using net-iplsearch in a host list.
5727 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5728 ACLs are no longer actually imposed (and a message to that effect is
5731 22. If a "gecos" field in a passwd entry contained escaped characters, in
5732 particular, if it contained a \" sequence, Exim got it wrong when building
5733 a From: or a Sender: header from that name. A second bug also caused
5734 incorrect handling when an unquoted " was present following a character
5735 that needed quoting.
5737 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5738 was not being matched caselessly.
5740 24. Arranged for all hyphens in the exim.8 source to be escaped with
5743 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5744 properly. Recipient callout cache records were still being keyed to include
5745 the sender, even when use_sender was set false. This led to far more
5746 callouts that were necessary. The sender is no longer included in the key
5747 when use_sender is false.
5749 26. Added "control = submission" modifier to ACLs.
5751 27. Added the ${base62d: operator to decode base 62 numbers.
5753 28. dnsdb lookups can now access SRV records.
5755 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5756 the configuration file.
5758 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5759 (-v) mode. This makes the output for a verbose queue run more intelligible.
5761 31. Added a use_postmaster feature to recipient callouts.
5763 32. Added the $body_zerocount variable, containing the number of binary zero
5764 bytes in the message body.
5766 33. The time of last modification of the "new" subdirectory is now used as the
5767 "mailbox time last read" when there is a quota error for a maildir
5770 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5772 35. Added +ignore_unknown as a special item in host lists.
5774 36. Code for decoding IPv6 addresses in host lists is now included, even if
5775 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5776 address was recognized as an IP address, but was then not correctly decoded
5777 into binary, causing unexpected and incorrect effects when compared with
5784 1. Very minor rewording of debugging text in manualroute to say "list of
5785 hosts" instead of "hostlist".
5787 2. If verify=header_syntax was set, and a header line with an unqualified
5788 address (no domain) and a large number of spaces between the end of the
5789 name and the colon was received, the reception process suffered a buffer
5790 overflow, and (when I tested it) crashed. This was caused by some obsolete
5791 code that should have been removed. The fix is to remove it!
5793 3. When running in the test harness, delay a bit after writing a bounce
5794 message to get a bit more predictability in the log output.
5796 4. Added a call to search_tidyup() just before forking a reception process. In
5797 theory, someone could use a lookup in the expansion of smtp_accept_max_
5798 per_host which, without the tidyup, could leave open a database connection.
5800 5. Added the variables $recipient_data and $sender_data which get set from a
5801 lookup success in an ACL "recipients" or "senders" condition, or a router
5802 "senders" option, similar to $domain_data and $local_part_data.
5804 6. Moved the writing of debug_print from before to after the "senders" test
5807 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5808 problems for message scanning, either using a data ACL, or using
5809 local_scan() because the Received: header was not generated till after they
5810 were called (in order to set the time as the time of reception completion).
5811 I have revised the way this works. The header is now generated after the
5812 body is received, but before the ACL or local_scan() are called. After they
5813 are run, the timestamp in the header is updated.
5819 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5820 before starting a queue runner without re-exec. This happened only when
5821 deliver_drop_privilege was set or when the Exim user was set to root. The
5822 effect of the bug was that timeouts during subsequent deliveries caused
5823 crashes instead of being properly handled. The handler is now left at its
5824 default (and expected) setting.
5826 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5827 message, again when deliver_drop_privilege is set or Exim is run as root.
5828 The bug described in (1) was not present in this case, but the tidying up
5829 of the other signals was missing. I have made the two cases consistent.
5831 3. The ignore_target_hosts setting on a manualroute router was being ignored
5832 for hosts that were looked up using the /MX notation.
5834 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5837 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5838 operated on the sender address. After changing the $sender_address to <>
5839 for the sender address verify, Exim was re-instated it as the original
5840 (before rewriting) address, but remembering that it had rewritten it, so it
5841 wasn't rewriting it again. This bug also had the effect of breaking the
5842 sender address verification caching when the sender address was rewritten.
5844 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5845 This has been changed so that if the ip literal address matches
5846 ignore_target_hosts, the router declines.
5848 7. Added expansion conditions match_domain, match_address, and match_local_
5849 part (NOT match_host).
5851 8. The placeholder for the Received: header didn't have a length field set.
5853 9. Added code to Exim itself and to exim_lock to test for a specific race
5854 condition that could lead to file corruption when using MBX delivery. The
5855 issue is with the lockfile that is created in /tmp. If this file is removed
5856 after a process has opened it but before that process has acquired a lock,
5857 there is the potential for a second process to recreate the file and also
5858 acquire a lock. This could lead to two Exim processes writing to the file
5859 at the same time. The added code performs the same test as UW imapd; it
5860 checks after acquiring the lock that its file descriptor still refers to
5861 the same named file.
5863 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5864 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5867 11. Added the smtp_active_hostname option. If used, this will typically be made
5868 to depend on the incoming interface address. Because $interface_address is
5869 not set up until the daemon has forked a reception process, error responses
5870 that can happen earlier (such as "too many connections") no longer contain
5873 12. If an expansion in a condition on a "warn" statement fails because a lookup
5874 defers, the "warn" statement is abandoned, and the next ACL statement is
5875 processed. Previously this caused the whole ACL to be aborted.
5877 13. Added the iplsearch lookup type.
5879 14. Added ident_timeout as a log selector.
5881 15. Added tls_certificate_verified as a log selector.
5883 16. Added a global option tls_require_ciphers (compare the smtp transport
5884 option of the same name). This controls incoming TLS connections.
5886 17. I finally figured out how to make tls_require_ciphers do a similar thing
5887 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5888 before starting the TLS session.
5890 18. Tabs are now shown as \t in -bP output.
5892 19. If the log selector return_path_on_delivery was set, Exim crashed when
5893 bouncing a message because it had too many Received: header lines.
5895 20. If two routers both had headers_remove settings, and the first one included
5896 a superfluous trailing colon, the final name in the first list and the
5897 first name in the second list were incorrectly joined into one item (with a
5898 colon in the middle).
5904 1. Added -C and -D options to the exinext utility, mainly to make it easier
5905 to include in the automated testing, but these could be helpful when
5906 multiple configurations are in use.
5908 2. The exinext utility was not formatting the output nicely when there was
5909 an alternate port involved in the retry record key, nor when there was a
5910 message id as well (for retries that were specific to a specific message
5911 and a specific host). It was also confused by IPv6 addresses, because of
5912 the additional colons they contain. I have fixed the IPv4 problem, and
5913 patched it up to do a reasonable job for IPv6.
5915 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5916 delivery, the log line now contains "pipelined" if PIPELINING was used.
5918 4. An SMTP transport process used to panic and die if the bind() call to set
5919 an explicit outgoing interface failed. This has been changed; it is now
5920 treated in the same way as a connect() failure.
5922 5. A reference to $sender_host_name in the part of a conditional expansion
5923 that was being skipped was still causing a DNS lookup. This no longer
5926 6. The def: expansion condition was not recognizing references to header lines
5927 that used bh_ and bheader_.
5929 7. Added the _cache feature to named lists.
5931 8. The code for checking quota_filecount in the appendfile transport was
5932 allowing one more file than it should have been.
5934 9. For compatibility with Sendmail, the command line option
5942 and sets the incoming protocol and host name (for trusted callers). The
5943 host name and its colon can be omitted when only the protocol is to be set.
5944 Note the Exim already has two private options, -pd and -ps, that refer to
5945 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5946 "s", but I don't think that's a major issue.
5948 10. A number of refactoring changes to the code, none of which should affect
5951 (a) The number of logging options was getting close to filling up the
5952 32-bit word that was used as a bit map. I have split them into two classes:
5953 those that are passed in the argument to log_write(), and those that are
5954 only ever tested independently outside of that function. These are now in
5955 separate 32-bit words, so there is plenty of room for expansion again.
5956 There is no change in the user interface or the logging behaviour.
5958 (b) When building, for example, log lines, the code previously used a
5959 macro that called string_cat() twice, in order to add two strings. This is
5960 not really sufficiently general. Furthermore, there was one instance where
5961 it was actually wrong because one of the argument was used twice, and in
5962 one call a function was used. (As it happened, calling the function twice
5963 did not affect the overall behaviour.) The macro has been replaced by a
5964 function that can join an arbitrary number of extra strings onto a growing
5967 (c) The code for expansion conditions now uses a table and a binary chop
5968 instead of a serial search (which was left over from when there were very
5969 few conditions). Also, it now recognizes conditions like "pam" even when
5970 the relevant support is not compiled in: a suitably worded error message is
5971 given if an attempt is made to use such a condition.
5973 11. Added ${time_interval:xxxxx}.
5975 12. A bug was causing one of the ddress fields not to be passed back correctly
5976 from remote delivery subprocesses. The field in question was not being
5977 subsequently used, so this caused to problems in practice.
5979 13. Added new log selectors queue_time and deliver_time.
5981 14. Might have fixed a bug in maildirsizefile handling that threw up
5982 "unexpected character" debug warnings, and recalculated the data
5983 unnecessarily. In any case, I expanded the warning message to give more
5986 15. Added the message "Restricted characters in address" to the statements in
5987 the default ACL that block characters like @ and % in local parts.
5989 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5990 Three changes have been made:
5992 (a) There was a serious bug; a negative response to MAIL caused the whole
5993 recipient domain to be cached as invalid, thereby blocking all messages
5994 to all local parts at the same domain, from all senders. This bug has
5995 been fixed. The domain is no longer cached after a negative response to
5996 MAIL if the sender used is not empty.
5998 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6001 (c) A new callout option, "use_sender" has been added for people who want
6002 the modified behaviour.
6008 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6011 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6012 indeed breaks things for older releases.
6014 3. Added additional logging to the case where there is a problem reading data
6015 from a filter that is running in a subprocess using a pipe, in order to
6016 try to track down a specific problem.
6018 4. Testing facility fudge: when running in the test harness and attempting
6019 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6020 getting "No route to host". Convert this to a timeout.
6022 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6025 6. Some OS don't have socklen_t but use size_t instead. This affects the
6026 fifth argument of getsockopt() amongst other things. This is now
6027 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6028 can be set for individual OS. I have set it for SunOS5, OSF1, and
6029 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6030 some earlier ones do not.
6032 7. Change 4.30/15 was not doing the test caselessly.
6034 8. The standard form for an IPv6 address literal was being rejected by address
6035 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6036 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6037 this, as well as the form without the "IPv6" on the front (but only when
6038 address literals are enabled, of course).
6040 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6042 10. Exim crashed if a message with an empty sender address specified by -f
6043 encountered a router with an errors_to setting. This could be provoked only
6044 by a command such as
6048 where an empty string was supplied; "<>" did not hit this bug.
6050 11. Installed PCRE release 4.5.
6052 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6053 remained set. It is now erased.
6055 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6056 times from message ids (which are base 36 rather than the normal 62).
6058 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6059 were being counted as actual protocol errors, and logged if the log
6060 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6061 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6062 and DATA following a set of rejected RCPTs do not count as protocol errors.
6063 In other words, Exim assumes they were pipelined, though this may not
6064 actually be the case. Of course, in all cases the client gets an
6065 appropriate error code.
6067 15. If a lookup fails in an ACL condition, a message about the failure may
6068 be available; it is used if testing the ACL cannot continue, because most
6069 such messages specify what the cause of the deferral is. However, some
6070 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6071 that caused an old message to be retained and used if a later statement
6072 caused a defer, replacing the real cause of the deferral.
6074 16. If an IP address had so many PTR records that the DNS lookup buffer
6075 was not large enough to hold them, Exim could crash while trying to process
6076 the truncated data. It now detects and logs this case.
6078 17. Further to 4.21/58, another change has been made: if (and only if) the
6079 first line of a message (the first header line) ends with CRLF, a bare LF
6080 in a subsequent header line has a space inserted after it, so as not to
6081 terminate the header.
6083 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6084 unnecessarily, used atoi() instead of strtol(), and didn't check the
6085 termination when getting file sizes from file names by regex.
6087 19. Completely re-implemented the support for maildirsize files, in the light
6088 of a number of problems with the previous contributed implementation
6089 (4.30/29). In particular:
6091 . If the quota is zero, the maildirsize file is maintained, but no quota is
6094 . If the maildir directory does not exist, it is created before any attempt
6095 to write a maildirsize file.
6097 . The quota value in the file is just a cache; if the quota is changed in
6098 the transport, the new value overrides.
6100 . A regular expression is available for excluding directories from the
6103 20. The autoreply transport checks the characters in options that define the
6104 message's headers; it allows continued headers, but it was checking with
6105 isspace() after an embedded newline instead of explicitly looking for a
6108 21. If all the "regular" hosts to which an address was routed had passed their
6109 expiry times, and had not reached their retry times, the address was
6110 bounced, even if fallback hosts were defined. Now Exim should go on to try
6113 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6114 equivalent code in the SMTP transport. Some hosts send humungous responses
6115 to HELO/EHLO, more than 1024 it seems.
6117 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6118 gives compiler warnings in some environments. I've now done it "properly",
6121 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6122 (because of problems with the built-in one) was declared to return uschar *
6123 instead of char *, causing compiler failure.
6125 25. Fixed a file descriptor leak when processing alias/forward files.
6127 26. Fixed a minor format string issue in dbfn.c.
6129 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6131 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6132 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6135 29. When a sender address is verified, it is cached, to save repeating the test
6136 when there is more than one recipient in a message. However, when the
6137 verification involves a callout, it is possible for different callout
6138 options to be set for different recipients. It is too complicated to keep
6139 track of this in the cache, so now Exim always runs a verification when a
6140 callout is required, relying on the callout cache for the optimization.
6141 The overhead is duplication of the address routing, but this should not be
6144 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6145 to be verified with callout=postmaster, and the main callout worked but the
6146 postmaster check failed, the verification correctly failed. However, if a
6147 subsequent RCPT command asked for sender verification *without* the
6148 postmaster check, incorrect caching caused this verification also to fail,
6151 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6152 it was not caching the DNS options (qualify_single, search_parents) that
6153 were used when the lookup failed. A subsequent lookup with different
6154 options therefore always gave the same answer, though there were cases
6155 where it should not have. (Example: a "domains = !$mx_any" option on a
6156 dnslookup router: the "domains" option is always processed without any
6157 widening, but the router might have qualify_single set.) Now Exim uses the
6158 cached value only when the same options are set.
6160 32. Added John Jetmore's "exipick" utility to the distribution.
6162 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6163 than a timeout (e.g. a certificate is required, and is not provided), an
6164 Exim server now closes the connection immediately. Previously it waited for
6165 the client to close - but if the client is SSL, it seems that they each
6166 wait for each other, leading to a delay before one of them times out.
6168 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6169 maintained 0.8.x compatibility because I don't think many are using it, and
6170 it is clearly obsolete.
6172 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6175 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6176 client certificate was expired. A simple patch fixes this, though I don't
6177 understand the full logic of why the verify callback is called multiple
6180 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6181 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6182 which causes problems with some clients (such as the Certicom SSL Plus
6183 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6184 disables the coutermeasure allowing Eudora to connect."
6186 38. Exim was not checking that a write() to a log file succeeded. This could
6187 lead to Bad Things if a log got too big, in particular if it hit a file
6188 size limit. Exim now panics and dies if it cannot write to a log file, just
6189 as it does if it cannot open a log file.
6191 39. Modified OS/Makefile-Linux so that it now contains
6193 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6195 The two -D definitions ensure that Exim is compiled with large file
6196 support, which makes it possible to handle log files that are bigger than
6199 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6200 instance) a domain was checked against a named list that involved a lookup,
6201 causing $domain_data to be set, then another domain was checked against the
6202 same list, then the first domain was re-checked, the value of $domain_data
6203 after the final check could be wrong. In particular, if the second check
6204 failed, it could be set empty. This bug probably also applied to
6207 41. The strip_trailing_dot option was not being applied to the address given
6208 with the -f command-line option.
6210 42. The code for reading a message's header from the spool was incrementing
6211 $received_count, but never initializing it. This meant that the value was
6212 incorrect (doubled) while delivering a message in the same process in which
6213 it was received. In the most common configuration of Exim, this never
6214 happens - a fresh exec is done - but it can happen when
6215 deliver_drop_privilege is set.
6217 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6218 it now includes up to 150 characters of the unexpected data in the log
6221 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6222 and building data strings. The size of both of these buffers was 10 000
6223 bytes - far larger than anybody would *ever* want, thought I. Needless to
6224 say, somebody hit the limit. I have increased the maximum line length to
6225 20 000 and the maximum data length of concatenated lines to 100 000. I have
6226 also fixed two bugs, because there was no checking on these buffers. Tsk,
6227 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6228 buffer is too small.
6230 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6231 lsearch lookups. Now it does.
6233 46. When parsing a route_list item in a manualroute router, a fixed-length
6234 buffer was used for the list of hosts. I made this 1024 bytes long,
6235 thinking that nobody would ever have a list of hosts that long. Wrong.
6236 Somebody had a whole pile of complicated expansion conditions, and the
6237 string was silently truncated, leading to an expansion error. It turns out
6238 that it is easier to change to an unlimited length (owing to other changes
6239 that have happened since this code was originally written) than to build
6240 structure for giving a limitation error. The length of the item that
6241 expands into the list of hosts is now unlimited.
6243 47. The lsearch lookup could not handle data where the length of text line was
6244 more than 4095 characters. Such lines were truncated, leading to shortened
6245 data being returned. It should now handle lines of any length.
6247 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6248 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6251 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6252 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6253 no longer happens in comment lines. A list of replacements is now placed
6254 at the head of all of the source files, except those whose only change is
6255 to replace PERL_COMMAND in the very first #! line.
6257 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6258 messages on the queue, with a bottom-up merge sort, using code contributed
6259 by Michael Haardt. This should make operations like -bp somewhat faster on
6260 large queues. It won't affect queue runners, except when queue_run_in_order
6263 51. Installed eximstats 1.31 in the distribution.
6265 52. Added support for SRV lookups to the dnslookup router.
6267 53. If an ACL referred to $message_body or $message_body_end, the value was not
6268 reset for any messages that followed in the same SMTP session.
6270 54. The store-handling optimization for building very long strings was not
6271 differentiating between the different store pools. I don't think this
6272 actually made any difference in practice, but I've tidied it.
6274 55. While running the routers to verify a sender address, $sender_address
6275 was still set to the sender address. This is wrong, because when routing to
6276 send a bounce to the sender, it would be empty. Therefore, I have changed
6277 it so that, while verifying a sender address, $sender_address is set to <>.
6278 (There is no change to what happens when verifying a recipient address.)
6280 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6281 target A or AAAA records (if not already returned) without resetting the
6282 qualify_single or search_parents options of the DNS resolver. These are
6283 inappropriate in this case because the targets of MX and SRV records must
6284 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6285 target that, when qualified, matched something in the local domain. These
6286 two options are now turned off when doing these lookups.
6288 57. It seems that at least some releases of Reiserfs (which does not have the
6289 concept of a fixed number of inodes) returns zero and not -1 for the
6290 number of available inodes. This interacted badly with check_spool_inodes,
6291 which assumed that -1 was the "no such thing" setting. What I have done is
6292 to check that the total number of inodes is greater than zero before doing
6293 the test of how many are available.
6295 58. When a "warn" ACL statement has a log_message modifier, the message is
6296 remembered, and not repeated. This is to avoid a lot of repetition when a
6297 message has many recipients that cause the same warning to be written.
6298 However, Exim was preserving the list of already written lines for an
6299 entire SMTP session, which doesn't seem right. The memory is now reset if a
6300 new message is started.
6302 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6303 debugging output unless log_rewrite was also set.
6305 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6306 of (int)(handle) when we know that handle contains (void *)(-1).
6308 61. The Exim daemon panic-logs an error return when it closes the incoming
6309 connection. However "connection reset by peer" seems to be common, and
6310 isn't really an error worthy of noting specially, so that particular error
6313 62. When Exim is trying to find all the local interfaces, it used to panic and
6314 die if the ioctl to get the interface flags failed. However, it seems that
6315 on at least one OS (Solaris 9) it is possible to have an interface that is
6316 included in the list of interfaces, but for which you get a failure error
6317 for this call. This happens when the interface is not "plumbed" into a
6318 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6319 failure of the "get flags" call assumes that the interface is down.
6321 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6322 makes life easier for people who are doing arithmetic on fields extracted
6323 from dates, where you often get leading zeros that should not be
6324 interpreted as octal.
6326 64. Added qualify_domain to the redirect router, to override the global
6329 65. If a pathologically long header line contained very many addresses (the
6330 report of this problem mentioned 10 000) and each of them was rewritten,
6331 Exim could use up a very large amount of memory. (It kept on making new
6332 copies of the header line as it rewrote, and never released the old ones.)
6333 At the expense of a bit more processing, the header rewriting function has
6334 been changed so that it no longer eats memory in this way.
6336 66. The generation of the Received: header has been moved from the time that a
6337 message starts to be received, to the time that it finishes. The timestamp
6338 in the Received: header should now be very close to that of the <= log
6339 line. There are two side-effects of this change:
6341 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6342 logged header lines no longer include the local Received: line, because
6343 it has not yet been created. The same applies to a copy of the message
6344 that is returned to a non-SMTP sender when a message is rejected.
6346 (b) When a filter file is tested using -bf, no additional Received: header
6347 is added to the test message. After some thought, I decided that this
6350 This change does not affect the value of $received_for. It is still set
6351 after address rewriting, but before local_scan() is called.
6353 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6355 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6356 gave an unhelpful panic error message, and a defer error. I have managed to
6357 change this behaviour so that it now rejects any supplied certificate,
6358 which seems right, as the list of acceptable certificates is empty.
6360 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6361 gave an unhelpful defer error. I have not managed to make this reject any
6362 supplied certificates, but the error message it gives is "no certificate
6363 supplied", which is not helpful.
6365 70. exigrep's output now also includes lines that are not associated with any
6366 message, but which match the given pattern. Implemented by a patch from
6367 Martin Sluka, which also tidied up the Perl a bit.
6369 71. Recipient callout verification, like sender verification, was using <> in
6370 the MAIL FROM command. This isn't really the right thing, since the actual
6371 sender may affect whether the remote host accepts the recipient or not. I
6372 have changed it to use the actual sender in the callout; this means that
6373 the cache record is now keyed on a recipient/sender pair, not just the
6374 recipient address. There doesn't seem to be a real danger of callout loops,
6375 since a callout by the remote host to check the sender would use <>.
6376 [SEE ABOVE: changed after hitting problems.]
6378 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6379 temporary errors. However, in the case of such a code being given after
6380 the end of a data transmission (i.e. after ".") Exim was failing to write
6381 a retry record for the message. (Yes, there was some broken host that was
6382 actually sending 8xx at this point.)
6384 73. An unknown lookup type in a host list could cause Exim to panic-die when
6385 the list was checked. (An example that provoked this was putting <; in the
6386 middle of a list instead of at the start.) If this happened during a DATA
6387 ACL check, a -D file could be left lying around. This kind of configuration
6388 error no longer causes Exim to die; instead it causes a defer error. The
6389 incident is still logged to the main and panic logs.
6391 74. Buglet left over from Exim 3 conversion. The message "too many messages
6392 in one connection" was written to the rejectlog but not the mainlog, except
6393 when address rewriting (yes!) was being logged.
6395 75. Added write_rejectlog option.
6397 76. When a system filter was run not as root (that is, when system_filter_user
6398 was set), the values of the $n variables were not being returned to the
6399 main process; thus, they were not subsequently available in the $sn
6402 77. Added +return_path_on_delivery log selector.
6404 78. A connection timeout was being treated differently from recipients deferred
6405 when testing hosts_max_try with a message that was older than the host's
6406 retry timeout. (The host should not be counted, thus allowing all hosts to
6407 be tried at least once before bouncing.) This may have been the cause of an
6408 occasionally reported bug whereby a message would remain on the queue
6409 longer than the retry timeout, but would be bounced if a delivery was
6410 forced. I say "may" because I never totally pinned down the problem;
6411 setting up timeout/retry tests is difficult. See also the next item.
6413 79. The ultimate address timeout was not being applied to errors that involved
6414 a combination of host plus message (for example, a timeout on a MAIL
6415 command). When an address resolved to a number of possible hosts, and they
6416 were not all tried for each delivery (e.g. because of hosts_max_try), a
6417 message could remain on the queue longer than the retry timeout.
6419 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6422 81. Fixed an obscure SMTP outgoing bug which required at least the following
6423 conditions: (a) there was another message waiting for the same server;
6424 (b) the server returned 5xx to all RCPT commands in the first message so
6425 that the message was not completed; (c) the server dropped the connection
6426 or gave a negative response to the RSET that Exim sends to abort the
6427 transaction. The observed case was a dropped connection after DATA that had
6428 been sent in pipelining mode. That is, the server had advertised PIPELINING
6429 but was not implementing it correctly. The effect of the bug was incorrect
6430 behaviour, such as trying another host, and this could lead to a crash.
6436 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6437 and daemon.c were passed as pointers to ints; they should have been
6438 pointers to socklen_t variables (which are typically unsigned ints).
6440 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6443 3. Fixed a really odd bug that affected only the testing scheme; patching a
6444 certain fixed string in the binary changed the value of another string that
6445 happened to be identical to the end of the original first string.
6447 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6448 name", it returns that address as the IP address. On some operating
6449 systems (e.g. Solaris), it also passes back the IP address string as the
6450 "host name". However, on others (e.g. Linux), it passes back an empty
6451 string. Exim wasn't checking for this, and was changing the host name to an
6452 empty string, assuming it had been canonicalized.
6454 5. Although rare, it is permitted to have more than one PTR record for a given
6455 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6456 all the names associated with an address, because they do in Solaris.
6457 However, it seems that they do not in Linux for data that comes from the
6458 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6459 I found this out when I moved to a new Linux workstation and tried to run
6460 the Exim test suite.
6462 To get round this problem I have changed the code so that it now does its
6463 own call to the DNS to look up PTR records when searching for a host name.
6464 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6465 addresses that are only in /etc/hosts are still found.
6467 This behaviour is, however, controlled by an option called host_lookup_
6468 order, which defaults to "bydns:byaddr". If people want to use the other
6469 order, or indeed, just use one or the other means of lookup, they can
6470 specify it in this variable.
6472 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6473 some operating systems, this comes back from gethostbyaddr() as an empty
6474 string, and this is what Exim used to test for. However, it seems that in
6475 other systems, "." is yielded. Exim now tests for this case too.
6477 7. The values of check_spool_space and check_log_space are now held internally
6478 as a number of kilobytes instead of an absolute number of bytes. If a
6479 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6480 kilobyte. This means that much larger values can be stored.
6482 8. Exim monitor: an attempt to get the action menu when not actually pointing
6483 at a message produces an empty menu entitled "No message selected". This
6484 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6485 no entries in it ("Shell widget menu has zero width and/or height"). So I
6486 have added a single, blank menu entry in this case.
6488 9. Added ${quote_local_part.
6490 10. MIME decoding is now applied to the contents of Subject: header lines when
6493 11. Now that a reference to $sender_host_address automatically causes a reverse
6494 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6495 host lookup before query-style lookups in lists that might use this
6496 variable. This has therefore been abolished, and the "net-" prefix is no
6497 longer necessary for query-style lookups.
6499 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6500 have been a typo for LFLAGS, so it has been changed.
6502 13. The install script calls Exim with "-C /dev/null" in order to find the
6503 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6504 to be output. However, since Exim outputs its version number before the
6505 error, it didn't break the script. It just looked ugly. I fixed this by
6506 always allowing "-C /dev/null" if the caller is root.
6508 14. Ignore overlarge ACL variable number when reading spool file - insurance
6509 against a later release with more variables having written the file.
6511 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6512 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6513 this, as well as the form without the "IPv6" on the front.
6515 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6516 OS/Makefile-Darwin file.
6518 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6519 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6521 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6522 variable. However, some OS use ints for inodes. I've added cast to long int
6523 to get rid of the compiler warning.
6525 19. I had forgotten to lock out "/../" in configuration file names when
6526 ALT_CONFIG_PREFIX was set.
6528 20. Routers used for verification do not need to specify transports. However,
6529 if such a router generated a host list, and callout was configured, Exim
6530 crashed, because it could not find a port number from the (non-existent)
6531 transport. It now assumes port 25 in this circumstance.
6533 21. Added the -t option to exigrep.
6535 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6536 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6537 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6538 them is set, LOOKUP_LSEARCH is forced.
6540 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6542 24. Added sender and host information to the "rejected by local_scan()" log
6543 line; previously there was no indication of these.
6545 25. Added .include_if_exists.
6547 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6548 Linux. It turns out that not all file systems support this. Apparently some
6549 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6550 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6551 not supported on the file descriptor, is now ignored when Exim is trying to
6552 sync a directory. This applies only to Linux.
6554 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6556 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6557 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6560 29. Added support for maildirsize files from supplied patch (modified a bit).
6562 30. The use of :fail: followed by an empty string could lead Exim to respond to
6563 sender verification failures with (e.g.):
6565 550 Verification failed for <xxx>
6566 550 Sender verify failed
6568 where the first response line was missing the '-' that indicates it is not
6569 the final line of the response.
6571 31. The loop for finding the name of the user that called Exim had a hardwired
6572 limit of 10; it now uses the value of finduser_retries, which is used for
6573 all other user lookups.
6575 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6578 33. Exim was neglecting to zero errno before one call of strtol() when
6579 expanding a string and expecting an integer value. On some systems this
6580 resulted in spurious "integer overflow" errors. Also, it was casting the
6581 result into an int without checking.
6583 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6584 did not work. The code looks as if it has *never* worked, though it appears
6585 to have been documented since at least release 1.62. I have made it work.
6587 35. The "timeout_DNS" error in retry rules, also documented since at least
6588 1.62, also never worked. As it isn't clear exactly what this means, and
6589 clearly it isn't a major issue, I have abolished the feature by treating it
6590 as "timeout", and writing a warning to the main and panic logs.
6592 36. The display of retry rules for -brt wasn't always showing the error code
6595 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6596 timeout_connect_A, timeout_connect_MX.
6598 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6599 to the empty sender.
6601 39. The daemon was not analysing the content of -oX till after it had closed
6602 stderr and disconnected from the controlling terminal. This meant that any
6603 syntax errors were only noted on the panic log, and the return code from
6604 the command was 0. By re-arranging the code a little, I've made the
6605 decoding happen first, so such errors now appear on stderr, and the return
6606 code is 1. However, the actual setting up of the sockets still happens in
6607 the disconnected process, so errors there are still only recorded on the
6610 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6611 connections (as happens on some IP stacks) was logged at start up time as
6612 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6613 it from "IPv6 and IPv4", which means that two separate sockets are being
6616 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6617 says whether AF_INET or AF_INET6 was passed as an argument.
6619 42. Exiwhat output was messed up when time zones were included in log
6622 43. Exiwhat now gives more information about the daemon's listening ports,
6623 and whether -tls-on-connect was used.
6625 44. The "port" option of the smtp transport is now expanded.
6627 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6628 silently ignored. Now an error message is written to the main and panic
6631 46. There's a new ACL modifier called "logwrite" which writes to a log file
6632 as soon as it is encountered.
6634 47. Added $local_user_uid and $local_user_gid at routing time.
6636 48. Exim crashed when trying to verify a sender address that was being
6639 49. Exim was recognizing only a space character after ".include". It now also
6640 recognizes a tab character.
6642 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6643 extracting the relevant information from the specification. The man page no
6644 longer contains scrambled data for the -d option, and I've added a section
6645 at the front about calling Exim under different names.
6647 51. Added "extra_headers" argument to the "mail" command in filter files.
6649 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6652 53. Installed eximstats 1.29.
6654 54. Added transport_filter_timeout as a generic transport option.
6656 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6657 Cc: header lines. This was required by RFC 822, but it not required by RFC
6660 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6661 incoming messages that did not have them. Now it does so only if the
6662 message originates locally, that is, if there is no associated remote host
6663 address. When Resent- header lines are present, this applies to the Resent-
6664 lines rather than the non-Resent- lines.
6666 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6667 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6669 58. Messages for configuration errors now include the name of the main
6670 configuration file - useful now that there may be more than one file in a
6671 list (.included file names were always shown).
6673 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6674 for those rare installations that do not start the daemon as root or run it
6675 setuid root. I've cut out the call to initgroups() if the daemon is not
6678 60. The Exim user and group can now be bound into the binary as text strings
6679 that are looked up at the start of Exim's processing.
6681 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6683 62. Added $mailstore_basename variable.
6685 63. Installed patch to sieve.c from Michael Haardt.
6687 64. When Exim failed to open the panic log after failing to open the main log,
6688 the original message it was trying to log was written to stderr and debug
6689 output, but if they were not available (the usual case in production), it
6690 was lost. Now it is written to syslog before the two lines that record the
6691 failures to open the logs.
6693 65. Users' Exim filters run in subprocesses under the user's uid. It is
6694 possible for a "deliver" command or an alias in a "personal" command to
6695 provoke an address rewrite. If logging of address rewriting is configured,
6696 this fails because the process is not running as root or exim. There may be
6697 a better way of dealing with this, but for the moment (because 4.30 needs
6698 to be released), I have disabled address rewrite logging when running a
6699 filter in a non-root, non-exim process.
6705 1. The buildconfig auxiliary program wasn't quoting the value set for
6706 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6707 not defined. This bug was masked in 4.22 by the effect that was fixed in
6710 2. Some messages that were rejected after a message id was allocated were
6711 shown as "incomplete" by exigrep. It no longer does this for messages that
6712 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6714 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6715 have allow_domain_literals set, the ID did not get logged in the <= line.
6716 Domain literals are now always recognized in Message-ID: header lines.
6718 4. The first argument for a ${extract expansion item is the key name or field
6719 number. Leading and trailing spaces in this item were not being ignored,
6720 causing some misleading effects.
6722 5. When deliver_drop_privilege was set, single queue runner processes started
6723 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6724 same command in the process it spins off) were not dropping privilege.
6726 6. When the daemon running as "exim" started a queue runner, it always
6727 re-executed Exim in the spun-off process. This is a waste of effort when
6728 deliver_drop_privilege is set. The new process now just calls the
6729 queue-runner function directly.
6735 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6738 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6739 ignored. Though the use of -oP was forcing the writing of a pid file, it
6740 was always written to the default place.
6742 3. If the message "no IP address found for host xxxx" is generated during
6743 incoming verification, it is now followed by identification of the incoming
6744 connection (so you can more easily find what provoked it).
6746 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6748 5. Added some features to "harden" Exim a bit more against certain attacks:
6750 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6751 be put in Local/Makefile. This is like the never_users runtime option,
6752 but it cannot be overridden. The default setting is "root".
6754 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6755 prefix string with which any file named in a -C command line option
6758 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6759 is retained for -C and -D only if the caller of Exim is root. Without
6760 it, the exim user may also use -C and -D and retain privilege.
6762 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6763 command line option is disabled.
6765 6. Macro names set by the -D option must start with an upper case letter, just
6766 like macro names defined in the configuration file.
6768 7. Added "dereference=" facility to LDAP.
6770 8. Two instances of the typo "uknown" in the source files are fixed.
6772 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6773 the Configure-Makefile script screwed up while processing it.
6775 10. Incorporated PCRE 4.4.
6777 11. The SMTP synchronization check was not operating right at the start of an
6778 SMTP session. For example, it could not catch a HELO sent before the client
6779 waited for the greeting. There is now a check for outstanding input at the
6780 point when the greeting is written. Because of the duplex, asynchronous
6781 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6782 way, but not yet received, when the check is performed.
6784 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6785 on TCP/IP sockets, because this apparently causes some broken clients to
6788 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6789 unchanged) from the Cygwin maintainer.
6791 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6792 was supported instead of testing for mbx. Effectively a typo.
6794 15. The spa authenticator server code was not checking that the input it
6795 received was valid base64.
6797 16. The debug output line for the "set" modifier in ACLs was not showing the
6798 name of the variable that was being set.
6800 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6802 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6803 lookup on its own. Something else was needed to trigger the lookup. For
6804 example, a match in host_lookup or the need for a host name in a host list.
6805 Now, if $sender_host_name is referenced and the host name has not yet been
6806 looked up, a lookup is performed. If the lookup fails, the variable remains
6807 empty, and $host_lookup_failed is set to "1".
6809 19. Added "eqi" as a case-independent comparison operator.
6811 20. The saslauthd authentication condition could segfault if neither service
6812 nor realm was specified.
6814 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6815 error message that was logged was misleading, and incoming SMTP
6816 connections were dropped. The message is now more accurate, and temporary
6817 errors are given to SMTP connections.
6819 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6820 (except RSET) with a 421 error, until QUIT is received. However, it was
6821 failing to send a response to QUIT.
6823 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6824 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6825 ACL is now run after the helo_try_verify_hosts code.
6827 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6828 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6829 case-independent, so other case variants are also recognized). Apparently
6830 some systems use these upper case variants.
6832 25. If more than two messages were waiting for the same host, and a transport
6833 filter was specified for the transport, Exim sent two messages over the
6834 same TCP/IP connection, and then failed with "socket operation on non-
6835 socket" when it tried to send the third.
6837 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6839 27. The extern definition of crypt16() in expand.c was not being excluded when
6840 the OS had its own crypt16() function.
6842 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6843 as a preferred synonym for return_size_limit, both as an option and as an
6846 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6848 30. Changed the default configuration ACL to relax the local part checking rule
6849 for addresses that are not in any local domains. For these addresses,
6850 slashes and pipe symbols are allowed within local parts, but the sequence
6851 /../ is explicitly forbidden.
6853 31. SPA server authentication was not clearing the challenge buffer before
6856 32. log_message in a "warn" ACL statement was writing to the reject log as
6857 well as to the main log, which contradicts the documentation and doesn't
6858 seem right (because no rejection is happening). So I have stopped it.
6860 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6861 However, I am unable to do any testing of this.
6863 34. Fixed an infelicity in the appendfile transport. When checking directories
6864 for a mailbox, to see if any needed to be created, it was accidentally
6865 using path names with one or more superfluous leading slashes; tracing
6866 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6868 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6869 contents are added to the log line that is written for every discarded
6870 recipient. (Previously a log_message setting was ignored.)
6872 36. The ${quote: operator now quotes the string if it is empty.
6874 37. The install script runs exim in order to find its version number. If for
6875 some reason other than non-existence or emptiness, which it checks, it
6876 could not run './exim', it was installing it with an empty version number,
6877 i.e. as "exim-". This error state is now caught, and the installation is
6880 38. An argument was missing from the function that creates an error message
6881 when Exim fails to connect to the socket for saslauthd authentication.
6882 This could cause Exim to crash, or give a corrupted message.
6884 39. Added isip, isip4, and isip6 to ${if conditions.
6886 40. The ACL variables $acl_xx are now saved with the message, and can be
6887 accessed later in routers, transports, and filters.
6889 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6890 strings in the file are not string-expanded.
6892 42. If a MAIL command specified a SIZE value that was too large to fit into an
6893 int variable, the check against message_size_limit failed. Such values are
6894 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6895 day this will have to be increased, but I don't think I want to be around
6896 when emails are that large.
6903 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6904 iconv() is not standard in FreeBSD.
6906 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6907 IPv6 enabled. The observed symptom was a segmentation fault on return from
6908 the function os_common_find_running_interfaces() in src/os.c.
6910 3. In the check_special_case() function in daemon.c I had used "errno" as an
6911 argument name, which causes warnings on some systems. This was basically a
6912 typo, since it was named "eno" in the comments!
6914 4. The code that waits for the clock to tick (at a resolution of some fraction
6915 of a second) so as to ensure message-id uniqueness was always waiting for
6916 at least one whole tick, when it could have waited for less. [This is
6917 almost certainly not relevant at current processor speeds, where it is
6918 unlikely to ever wait at all. But we try to future-proof.]
6920 5. The function that sleeps for a time interval that includes fractions of a
6921 second contained a race. It did not block SIGALRM between setting the
6922 timer, and suspending (a couple of lines later). If the interval was short
6923 and the sigsuspend() was delayed until after it had expired, the suspension
6924 never ended. On busy systems this could lead to processes getting stuck for
6927 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6928 process, before it forks any delivery processes. The open lookup caching
6929 mechanism meant that the open file or database connection was passed into
6930 the delivery process. The problem was that delivery processes always tidy
6931 up cached lookup data. This could cause a problem for the next delivery
6932 process started by the queue runner, because the external queue runner
6933 process does not know about the closure. So the next delivery process
6934 still has data in the lookup cache. In the case of a file lookup, there was
6935 no problem because closing a file descriptor in a subprocess doesn't affect
6936 the parent. However, if the lookup was caching a connection to a database,
6937 the connection was closed, and the second delivery process was likely to
6938 see errors such as "PGSQL: query failed: server closed the connection
6939 unexpectedly". The problem has been fixed by closing all cached lookups
6940 in a queue runner before running a delivery process.
6942 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6943 seem to have the "const" qualifier which it has on other OS. I've
6946 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6947 *of the same type* (client or server) with the same public name that an
6948 error should be diagnosed.
6950 9. When Exim looked up a host name for an IP address, but failed to find the
6951 original IP address when looking up the host name (a safety check), it
6952 output the message "<ip address> does not match any IP for NULL", which was
6953 confusing, to say the least. The bug was that the host name should have
6954 appeared instead of "NULL".
6956 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6957 user that is built into the binary, and the -C or -D options is used, root
6958 privilege is dropped before the configuration file is read. In addition,
6959 logging is switched to stderr instead of the normal log files. If the
6960 configuration then re-defines the Exim user, the unprivileged environment
6961 is probably not what is expected, so Exim logs a panic warning message (but
6964 However, if deliver_drop_privilege is set, the unprivileged state may well
6965 be exactly what is intended, so the warning has been cut out in that case,
6966 and Exim is allowed to try to write to its normal log files.
6972 1. smtp_return_error_details was not giving details for temporary sender
6973 or receiver verification errors.
6975 2. Diagnose a configuration error if two authenticators have the same public
6978 3. Exim used not to create the message log file for a message until the first
6979 delivery attempt. This could be confusing when incoming messages were held
6980 for policy or load reasons. The message log file is now created at the time
6981 the message is received, and an initial "Received" line is written to it.
6983 4. The automatically generated man page for command line options had a minor
6984 bug that caused no ill effects; however, a more serious problem was that
6985 the procedure for building the man page automatically didn't always
6986 operate. Consequently, release 4.20 contains an out-of-date version. This
6987 shouldn't happen again.
6989 5. When building Exim with embedded Perl support, the script that builds the
6990 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6991 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6993 6. The freeze_tell option was not being used for messages that were frozen on
6994 arrival, either by an ACL or by local_scan().
6996 7. Added the smtp_incomplete_transaction log selector.
6998 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6999 was accepting AUTH without a new EHLO.
7001 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7002 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7005 10. Logging of TCP/IP connections (when configured) now happens in the main
7006 daemon process instead of the child process, so that the TCP/IP connection
7007 count is more accurate (but it can never be perfect).
7009 11. The use of "drop" in a nested ACL was not being handled correctly in the
7010 outer ACL. Now, if condition failure induced by the nested "drop" causes
7011 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7012 or "require"), the connection is dropped.
7014 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7015 that yield "discard" can now be used with an "accept" or a "discard" verb,
7016 but an error is generated for any others (because I can't see a useful way
7017 to define what should happen).
7019 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7020 are now processed in the same way as lines in the Exim configuration file.
7021 In particular, continuation lines are supported.
7023 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7025 15. Added -ti meaning -t -i.
7027 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7028 domains, and warn by logging if others are found.
7030 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7031 returned from GIFCONF when Exim is trying to find the list of interfaces on
7032 a host. The code in os.c has been modified to copy each ifreq to an aligned
7033 structure in all cases.
7035 Also, in some cases, the returned ifreq's were being copied to a 'struct
7036 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7037 means the last couple of bytes of an IPv6 address could be chopped if the
7038 ifreq contained only a normal sockaddr (14 bytes storage).
7040 18. Named domain lists were not supported in the hosts_treat_as_local option.
7041 An entry such as +xxxx was not recognized, and was treated as a literal
7044 19. Ensure that header lines added by a DATA ACL are included in the reject log
7045 if the ACL subsequently rejects the message.
7047 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7048 MD5 (which is deprecated).
7050 21. When testing a filter file using -bf, Exim was writing a message when it
7051 took the sender from a "From " line in the message, but it was not doing so
7052 when it took $return_path from a Return-Path: header line. It now does.
7054 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7055 with a valid header line field name (a series of printing characters
7056 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7058 23. Changed "disc" in the source to "disk" to conform to the documentation and
7059 the book and for uniformity.
7061 24. Ignore Sendmail's -Ooption=value command line item.
7063 25. When execve() failed while trying to run a command in a pipe transport,
7064 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7065 could be confused with a return value of 69 from the command itself. This
7066 has been changed to 127, the value the shell returns if it is asked to run
7067 a non-existent command. The wording for the related log line suggests a
7068 non-existent command as the problem.
7070 26. If received_header_text expands to an empty string, do not add a Received:
7071 header line to the message. (Well, it adds a token one on the spool, but
7072 marks it "old" so that it doesn't get used or transmitted.)
7074 27. Installed eximstats 1.28 (addition of -nt option).
7076 28. There was no check for failure on the call to getsockname() in the daemon
7077 code. This can fail if there is a shortage of resources on the system, with
7078 ENOMEM, for example. A temporary error is now given on failure.
7080 29. Contrary to the C standard, it seems that in some environments, the
7081 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7082 program. Exim now does this explicitly; it affects the formatting of
7083 timestamps using strftime().
7085 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7086 complaints. I've now initialized all the variables, to avoid this.
7088 32. Header lines added by a system filter were not being "seen" during
7089 transport-time rewrites.
7091 33. The info_callback() function passed to OpenSSL is set up with type void
7092 (*)(SSL *, int, int), as described somewhere. However, when calling the
7093 function (actually a macro) that sets it up, the type void(*)() is
7094 expected. I've put in a cast to prevent warnings from picky compilers.
7096 34. If a DNS black list lookup found a CNAME record, but there were no A
7097 records associated with the domain it pointed at, Exim crashed.
7099 35. If a DNS black list lookup returned more than one A record, Exim ignored
7100 all but the first. It now scans all returned addresses if a particular IP
7101 value is being sought. In this situation, the contents of the
7102 $dnslist_value variable are a list of all the addresses, separated by a
7105 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7106 to accept a host name and all its aliases if the forward lookup for any of
7107 them yielded the IP address of the incoming connection. Now it accepts only
7108 those names whose forward lookup yields the correct IP address. Any other
7109 names are discarded. This closes a loophole whereby a rogue DNS
7110 administrator could create reverse DNS records to break through a
7111 wildcarded host restriction in an ACL.
7113 37. If a user filter or a system filter that ran in a subprocess used any of
7114 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7115 the wrong values were passed to the pipe command ($thisaddress had the
7116 value of $0, $0 had the value of $1, etc). This bug was introduced by
7117 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7119 38. Improved the line breaking for long SMTP error messages from ACLs.
7120 Previously, if there was no break point between 40 and 75 characters, Exim
7121 left the rest of the message alone. Two changes have been made: (a) I've
7122 reduced the minimum length to 35 characters; (b) if it can't find a break
7123 point between 35 and 75 characters, it looks ahead and uses the first one
7124 that it finds. This may give the occasional overlong line, but at least the
7125 remaining text gets split now.
7127 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7128 file descriptors might be low, and that setting 1000 would always raise it.
7129 It turns out that in some environments, the limit is already over 1000 and
7130 that lowering it causes trouble. So now Exim takes care not to decrease it.
7132 40. When delivering a message, the value of $return_path is set to $sender_
7133 address at the start of routing (routers may change the value). By an
7134 oversight, this default was not being set up when an address was tested by
7135 -bt or -bv, which affected the outcome if any router or filter referred to
7138 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7139 log only when "message" or "log_message" are set. However, if one of the
7140 conditions was an address verification, or a call to a nested ACL, the
7141 messages generated by the underlying test were being passed through. This
7142 no longer happens. The underlying message is available in $acl_verify_
7143 message for both "message" and "log_message" expansions, so it can be
7144 passed through if needed.
7146 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7147 new expansion $bh_ to give the encoded byte string without charset
7148 translation. Translation happens only if iconv() is available; HAVE_ICONV
7149 indicates this at build time. HEADERS_CHARSET gives the charset to
7150 translate to; headers_charset can change it in the configuration, and
7151 "headers charset" can change it in an individual filter file.
7153 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7154 that creates RFC 2047 encoded "words" labels them as that charset instead
7155 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7156 expansion operator; (ii) when Exim creates a From: line for a local
7157 message; (iii) when a header line is rewritten to include a "phrase" part.
7159 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7160 buggy, causing it to skip the first lines of messages whose message ID
7161 ended in 'D'. This would not have bitten before Exim release 4.14, because
7162 message IDs were unlikely to end in 'D' before then. The effect was to have
7163 incorrect size information for certain domains.
7165 45. #include "config.h" was missing at the start of the crypt16.c module. This
7166 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7169 46. If there was a timeout during a "random" callout check, Exim treated it as
7170 a failure of the random address, and carried on sending RSET and the real
7171 address. If the delay was just some slowness somewhere, the response to the
7172 original RCPT would be taken as a response to RSET and so on, causing
7173 mayhem of various kinds.
7175 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7176 when I implemented it. It didn't allow for the fact that some option values
7177 may legitimately be negative (e.g. size_addition), and it didn't even do
7178 the right test for positive values.
7180 48. Domain names in DNS records are case-independent. Exim always looks them up
7181 in lower case. Some resolvers return domain names in exactly the case they
7182 appear in the zone file, that is, they may contain uppercase letters. Not
7183 all resolvers do this - some return always lower case. Exim was treating a
7184 change of case by a resolver as a change of domain, similar to a widening
7185 of a domain abbreviation. This triggered its re-routing code and so it was
7186 trying to route what was effectively the same domain again. This normally
7187 caused routing to fail (because the router wouldn't handle the domain
7188 twice). Now Exim checks for this case specially, and just changes the
7189 casing of the domain that it ultimately uses when it transmits the message
7192 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7195 50. If a filter generated a file delivery with a non-absolute name (possible if
7196 no home directory exists for the router), the forbid_file option was not
7199 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7200 the existing equality matching.
7202 52. Exim was using ints instead of ino_t variables in some places where it was
7203 dealing with inode numbers.
7205 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7206 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7207 called TMPDIR, and if it finds it is different, it changes its value.
7209 54. The smtp_printf() function is now made available to local_scan() so
7210 additional output lines can be written before returning. There is also an
7211 smtp_fflush() function to enable the detection of a dropped connection.
7212 The variables smtp_input and smtp_batched_input are exported to
7215 55. Changed the default runtime configuration: the message "Unknown user"
7216 has been removed from the ACL, and instead placed on the localuser router,
7217 using the cannot_route_message feature. This means that any verification
7218 failures that generate their own messages won't get overridden. Similarly,
7219 the "Unrouteable address" message that was in the ACL for unverifiable
7220 relay addresses has also been removed.
7222 56. Added hosts_avoid_esmtp to the smtp transport.
7224 57. The exicyclog script was not checking for the esoteric option
7225 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7226 will work only if exicyclog is run under the appropriate euid.
7228 58. Following a discussion on the list, the rules by which Exim recognises line
7229 endings on incoming messages have been changed. The -dropcr and drop_cr
7230 options are now no-ops, retained only for backwards compatibility. The
7231 following line terminators are recognized: LF CRLF CR. However, special
7232 processing applies to CR:
7234 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7235 nor a local message in the state where . is a terminator.
7237 (ii) If a bare CR is encountered in a header line, an extra space is added
7238 after the line terminator so as not to end the header. The reasoning
7239 behind this is that bare CRs in header lines are most likely either
7240 to be mistakes, or people trying to play silly games.
7242 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7243 was being incorrectly given as 18 bytes larger than it should have been.
7244 This is a VOB (very old bug).
7246 60. This may never have affected anything current, but just in case it has:
7247 When the local host is found other than at the start of a list of hosts,
7248 the local host, those with the same MX, and any that follow, are discarded.
7249 When the list in question was part of a longer list of hosts, the following
7250 hosts (not currently being processed) were also being discarded. This no
7251 longer happens. I'm not sure if this situation could ever has previously
7254 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7257 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7258 header. This is something that is recommended in a new Internet Draft, and
7259 is something that is documented as being done by Sendmail. There are two
7260 possible values. For messages generated by the autoreply transport, Exim
7263 Auto-Submitted: auto-replied
7265 whereas for all other generated messages (e.g. bounces) it adds
7267 Auto-Submitted: auto-generated
7269 63. The "personal" condition in filters now includes a test for the
7270 Auto-Submitted: header. If it contains the string "auto-" the message it
7271 not considered personal.
7273 64. Added rcpt_include_affixes as a generic transport option.
7275 65. Added queue_only_override (default true).
7277 66. Added the syslog_duplication option.
7279 67. If what should have been the first header line of a message consisted of
7280 a space followed by a colon, Exim was mis-interpreting it as a header line.
7281 It isn't of course - it is syntactically invalid and should therefore be
7282 treated as the start of the message body. The misbehaviour could have
7283 caused a number of strange effects, including loss of data in subsequent
7284 header lines, and spool format errors.
7286 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7287 client host had authenticated. This control can now be exercised by an ACL
7288 for more flexibility.
7290 69. By default, callouts do not happen when testing with -bh. There is now a
7291 variant, -bhc, which does actually run the callout code, including
7292 consulting and updating the callout cache.
7294 70. Added support for saslauthd authentication, courtesy of Alexander
7297 71. If statvfs() failed on the spool or log directories while checking their
7298 size for availability, Exim confusingly gave the error "space shortage".
7299 Furthermore, in debugging mode it crashed with a floating point exception.
7300 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7301 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7302 serious problem, Exim now writes to the main and panic logs when this
7303 happens, with details of the failure. It then refuses to accept the
7304 incoming message, giving the message "spool directory problem" or "log
7305 directory problem" with a 421 code for SMTP messages.
7307 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7308 0, 1, and 2 exist, because some OS complain for execs without them (see
7309 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7310 descriptors. However, the code omitted to check that the open succeeded,
7311 causing mysterious errors if for some reason the permissions on /dev/null
7312 got screwed. Now Exim writes a message to the main and panic logs, and
7313 bombs out if it can't open /dev/null.
7315 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7316 interact so that it is all more flexible. It is supposed to remain
7317 backwards compatible. Also added extra_local_interfaces.
7319 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7320 to bomb out with an assertion failure - to the client this appears as a
7321 connection drop. This problem occurs in the part of the code that was taken
7322 from the Samba project. Fortunately, the assertion is in a very simple
7323 function, so I have fixed this by reproducing the function inline in the
7324 one place where it is called, and arranging for authentication to fail
7325 instead of killing the process with assert().
7327 75. The SPA client code was not working when the server requested OEM rather
7328 than Unicode encoding.
7330 76. Added code to make require_files with a specific uid setting more usable in
7331 the case where statting the file as root fails - usually a non-root-mounted
7332 NFS file system. When this happens and the failure is EACCES, Exim now
7333 forks a subprocess and does the per-uid checking as the relevant uid.
7335 77. Added process_log_path.
7337 78. If log_file_path was not explicitly set, a setting of check_log_space or
7338 check_log_inodes was ignored.
7340 79. If a space check for the spool or log partitions fails, the incident is now
7341 logged. Of course, in the latter case the data may get lost...
7343 80. Added the %p formatting code to string_format() so that it can be used to
7344 print addresses in debug_print(). Adjusted all the address printing in the
7345 debugging in store.c to use %p rather than %d.
7347 81. There was a concern that a line of code in smtp_in.c could overflow a
7348 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7349 initially expressed, the concern was not well-founded, because trailing
7350 spaces are removed early. However, if the trailing spaces were followed by
7351 a NULL, they did not get removed, so the overflow was possible. Two fixes
7354 (a) I re-wrote the offending code in a cleaner fashion.
7355 (b) If an incoming SMTP command contains a NULL character, it is rejected
7358 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7359 runs initgroups(), so that if the Exim user is in any additional groups,
7360 they will be used during message reception.
7366 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob