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.
75 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
76 When redis_servers is set to a list of > 1 element, and the Redis servers
77 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
78 case of MOVED into a DEFER case instead, thus moving the query onto the
79 next server in the list. For a cluster of N elements, all N servers must
80 be defined in redis_servers.
82 GF/02 Catch and remove uninitialized value warning in exiqsumm
83 Check for existence of @ARGV before looking at $ARGV[0]
85 JH/01 Replace the store_release() internal interface with store_newblock(),
86 which internalises the check required to safely use the old one, plus
87 the allocate and data copy operations duplicated in both (!) of the
90 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
91 modifier. This matches the restriction on the commandline.
93 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
94 Previously only the last row was returned.
96 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
97 we assumed that tags in the header were well-formed, and parsed the
98 element content after inspecting only the first char of the tag.
99 Assumptions at that stage could crash the receive process on malformed
102 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
103 While running the DKIM ACL we operate on the Permanent memory pool so that
104 variables created with "set" persist to the DATA ACL. Also (at any time)
105 DNS lookups that fail create cache records using the Permanent pool. But
106 expansions release any allocations made on the current pool - so a dnsdb
107 lookup expansion done in the DKIM ACL releases the memory used for the
108 DNS negative-cache, and bad things result. Solution is to switch to the
109 Main pool for expansions.
110 While we're in that code, add checks on the DNS cache during store_reset,
111 active in the testsuite.
112 Problem spotted, and debugging aided, by Wolfgang Breyha.
114 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
115 When none of the hosts presented to a transport match an already-open
116 connection, close it and proceed with the list. Previously we would
117 queue the message. Spotted by Lena with Yahoo, probably involving
120 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
121 Previously a spurious "250 OK id=" response was appended to the proper
124 JH/08 The "support for" informational output now, which built with Content
125 Scanning support, has a line for the malware scanner interfaces compiled
126 in. Interface can be individually included or not at build time.
128 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
129 by the template makefile "src/EDITME". The "STREAM" support for an older
130 ClamAV interface method is removed.
132 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
133 rows affected is given instead).
135 JH/11 The runtime Berkeley DB library version is now additionally output by
136 "exim -d -bV". Previously only the compile-time version was shown.
138 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
139 SMTP connection. Previously, when one had more receipients than the
140 first, an abortive onward connection was made. Move to full support for
141 multiple onward connections in sequence, handling cutthrough connection
142 for all multi-message initiating connections.
144 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
145 routers. Previously, a multi-recipient message would fail to match the
146 onward-connection opened for the first recipient, and cause its closure.
148 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
149 a timeout on read on a GnuTLS initiating connection, resulting in the
150 initiating connection being dropped. This mattered most when the callout
151 was marked defer_ok. Fix to keep the two timeout-detection methods
154 JH/15 Relax results from ACL control request to enable cutthrough, in
155 unsupported situations, from error to silently (except under debug)
156 ignoring. This covers use with PRDR, frozen messages, queue-only and
159 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
161 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
162 metadata, resulting in a crash in free().
164 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
165 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
166 Broken also in d185889f4, with init system revamp.
168 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
169 Previously we abruptly closed the connection after reading a malware-
170 found indication; now we go on to read the "scan ok" response line,
173 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
174 ACL. Previously, a crash would result.
176 JH/19 Speed up macro lookups during configuration file read, by skipping non-
177 macro text after a replacement (previously it was only once per line) and
178 by skipping builtin macros when searching for an uppercase lead character.
180 JH/20 DANE support moved from Experimental to mainline. The Makefile control
181 for the build is renamed.
183 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
184 was allocated for every new TLS startup, meaning one per message. Fix
185 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
187 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
188 reported the original. Fix to report (as far as possible) the ACL
189 result replacing the original.
191 JH/23 Fix memory leak during multi-message connections using STARTTLS under
192 OpenSSL. Certificate information is loaded for every new TLS startup,
193 and the resources needed to be freed.
195 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
197 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
200 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
201 DATA response info to the (existing) per-recipient response info for
202 the "C=" log element. It can have useful tracking info from the
203 destination system. Patch from Simon Arlott.
205 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
206 length value. Previously this would segfault.
208 HS/02 Support Avast multiline protoocol, this allows passing flags to
209 newer versions of the scanner.
211 JH/28 Ensure that variables possibly set during message acceptance are marked
212 dead before release of memory in the daemon loop. This stops complaints
213 about them when the debug_store option is enabled. Discovered specifically
214 for sender_rate_period, but applies to a whole set of variables.
215 Do the same for the queue-runner and queue-list loops, for variables set
216 from spool message files. Do the same for the SMTP per-message loop, for
217 certain variables indirectly set in ACL operations.
219 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
220 as a multi-recipient message from a mailinglist manager). The coding had
221 an arbitrary cutoff number of characters while checking for more input;
222 enforced by writing a NUL into the buffer. This corrupted long / fast
223 input. The problem was exposed more widely when more pipelineing of SMTP
224 responses was introduced, and one Exim system was feeding another.
225 The symptom is log complaints of SMTP syntax error (NUL chars) on the
226 receiving system, and refused recipients seen by the sending system
227 (propating to people being dropped from mailing lists).
228 Discovered and pinpointed by David Carter.
230 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
231 replaced by the ${authresults } expansion.
233 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
235 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
236 allows proper process termination in container environments.
238 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
239 Previously the "final dot" had a newline after it; ensure it is CR,LF.
241 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
242 and "err_perm", deprecated since 4.83 when the RFC-defined words
243 "temperror" and "permerror" were introduced.
245 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
246 transport-filters or DKIM-signing. The restriction was lost in the
247 consolidation of verify-callout and delivery SMTP handling.
248 Extend the restriction to also cover ARC-signing.
250 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
251 in defer=pass mode supply a 450 to the initiator. Previously the message
254 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
255 tls_require_ciphers is used as before.
257 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
258 "pass_unscanned". Only tmpfails from the scanner are written to
259 the paniclog, as they may require admin intervention (permission
260 denied, license issues). Other scanner errors (like decompression
261 bombs) do not cause a paniclog entry.
263 JH/36 Fix reinitialisation of DKIM logging variable between messages.
264 Previously it was possible to log spurious information in receive log
267 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
268 triggered odd behaviour from Outlook Express clients.
270 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
273 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
274 since the IETF WG has not yet settled on that versus the original
275 "bare" representation.
277 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
278 Previously the millisecond value corrupted the output.
279 Fix also for syslog_pid=no and log_selector +pid, for which the pid
280 corrupted the output.
286 JH/01 Rework error string handling in TLS interface so that the caller in
287 more cases is responsible for logging. This permits library-sourced
288 string to be attached to addresses during delivery, and collapses
289 pairs of long lines into single ones.
291 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
292 during configuration. Wildcards are allowed and expanded.
294 JH/02 Rework error string handling in DKIM to pass more info back to callers.
295 This permits better logging.
297 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
298 do not close it down and have the child transport start it up again on
299 the passed-on TCP connection. Instead, proxy the child (and any
300 subsequent ones) for TLS via a unix-domain socket channel. Logging is
301 affected: the continued delivery log lines do not have any DNSSEC, TLS
302 Certificate or OCSP information. TLS cipher information is still logged.
304 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
305 identical IP addresses on different listening ports. Will also affect
308 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
309 add noisy ifdef guards to special-case this sillyness.
310 Patch from Bernd Kuhls.
312 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
313 than 255 are no longer allowed.
315 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
316 Disable the session-cache too, which might reduce our load. Since we
317 currrectly use a new context for every connection, both as server and
318 client, there is no benefit for these.
319 GnuTLS appears to not support tickets server-side by default (we don't
320 call gnutls_session_ticket_enable_server()) but client side is enabled
321 by default on recent versions (3.1.3 +) unless the PFS priority string
324 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
325 <https://reproducible-builds.org/specs/source-date-epoch/>.
327 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
328 the check for any unsuccessful recipients did not notice the limit, and
329 erroneously found still-pending ones.
331 JH/08 Pipeline CHUNKING command and data together, on kernels that support
332 MSG_MORE. Only in-clear (not on TLS connections).
334 JH/09 Avoid using a temporary file during transport using dkim. Unless a
335 transport-filter is involved we can buffer the headers in memory for
336 creating the signature, and read the spool data file once for the
337 signature and again for transmission.
339 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
340 4.77 as the kernel support then wasn't solid, having issues in 64bit
341 mode. Now, it's been long enough. Add support for FreeBSD also.
343 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
344 case where the routing stage had gathered several addresses to send to
345 a host before calling the transport for the first, we previously failed
346 to close down TLS in the old transport process before passing the TCP
347 connection to the new process. The new one sent a STARTTLS command
348 which naturally failed, giving a failed delivery and bloating the retry
349 database. Investigation and fix prototype from Wolfgang Breyha.
351 JH/12 Fix check on SMTP command input synchronisation. Previously there were
352 false-negatives in the check that the sender had not preempted a response
353 or prompt from Exim (running as a server), due to that code's lack of
354 awareness of the SMTP input buffering.
356 PP/04 Add commandline_checks_require_admin option.
357 Exim drops privileges sanely, various checks such as -be aren't a
358 security problem, as long as you trust local users with access to their
359 own account. When invoked by services which pass untrusted data to
360 Exim, this might be an issue. Set this option in main configuration
361 AND make fixes to the calling application, such as using `--` to stop
364 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
365 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
366 responses to those, into a single TLS record each way (this usually means
367 a single packet). As a side issue, smtp_enforce_sync now works on TLS
370 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
371 affects you only if you're dancing at the edge of the param size limits.
372 If you are, and this message makes sense to you, then: raise the
373 configured limit or use OpenSSL 1.1. Nothing we can do for older
376 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
377 cmdline element to get the documented default one. Previously it was
380 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
381 are unsuable for non-message-specific reasons - eg. connection timeout,
382 banner-time rejection.
384 JH/16 Fix logging of delivery remote port, when specified by router, under
387 PP/06 Repair manualroute's ability to take options in any order, even if one
388 is the name of a transport.
391 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
393 JH/17 Change the list-building routines interface to use the expanding-string
394 triplet model, for better allocation and copying behaviour.
396 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
397 Previously it was constructed the first time a possibly-matching string
398 was met in the configuration file input during startup; now it is done
401 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
402 compatible one, to avoid the (poorly documented) possibility of a config
403 file in the working directory redirecting the DB files, possibly correpting
404 some existing file. CVE-2017-10140 assigned for BDB.
406 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
407 cache-hot. Previously, although the result was properly cached, the
408 initial verify call returned a defer.
410 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
411 the main verify for receipient in uncached-mode.
413 JH/22 Retire historical build files to an "unsupported" subdir. These are
414 defined as "ones for which we have no current evidence of testing".
416 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
417 if present. Previously it was ignored.
419 JH/24 Start using specified-initialisers in C structure init coding. This is
420 a C99 feature (it's 2017, so now considered safe).
422 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
423 if was a fixed-sized field and bitmask ops via macros; it is now more
426 PP/07 GitHub PR 56: Apply MariaDB build fix.
427 Patch provided by Jaroslav Škarvada.
429 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
430 during Coverity cleanups [4.87 JH/47]
431 Diagnosis and fix provided by Michael Fischer v. Mollard.
433 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
434 the right size to place the terminating semicolon on its own folded
435 line, the header hash was calculated to an incorrect value thanks to
436 the (relaxed) space the fold became.
438 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
439 and confused the parent.
441 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
442 which could crash as a result. This could lead to undeliverable messages.
444 JH/28 Logging: "next input sent too soon" now shows where input was truncated
447 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
448 matters on fast-turnover and PID-randomising systems, which were getting
449 out-of-order delivery.
451 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
452 a possibly-overlapping copy. The symptom was that "Remote host closed
453 connection in response to HELO" was logged instead of the actual 4xx
456 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
457 Previously only that bufferd was discarded, resulting in SYMTP command
460 JH/32 DKIM: when a message has multiple signatures matching an identity given
461 in dkim_verify_signers, run the dkim acl once for each. Previously only
462 one run was done. Bug 2189.
464 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
465 "panic the current process" to "deliberately defer". The panic log is
466 still written with the problem list name; the mail and reject logs now
467 get a temp-reject line for the message that was being handled, saying
468 something like "domains check lookup or other defer". The SMTP 451
469 message is still "Temporary local problem".
471 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
472 A crafted sequence of BDAT commands could result in in-use memory beeing
473 freed. CVE-2017-16943.
475 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
476 from SMTP input. Previously it was always done; now only done for DATA
477 and not BDAT commands. CVE-2017-16944.
479 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
480 to the message (such as an overlong header line). Previously this was
481 not done and we did not exit BDAT mode. Followon from the previous item
482 though a different problem.
488 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
489 than -2003 did; needs libidn2 in addition to libidn.
491 JH/02 The path option on a pipe transport is now expanded before use.
493 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
494 Patch provided by "Björn", documentation fix added too.
496 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
497 missing a wire-to-host endian conversion.
499 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
500 close after a BDAT command line could be taken as a following command,
501 giving a synch failure. Fix by only checking for synch immediately
502 before acknowledging the chunk.
504 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
505 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
506 macro. Patches provided by Josh Soref.
508 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
509 Previously we did not; the RFC seems ambiguous and VRFY is not listed
510 by IANA as a service extension. However, John Klensin suggests that we
513 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
514 the dkim code may be unix-mode line endings rather than smtp wire-format
515 CRLF, so prepend a CR to any bare LF.
517 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
518 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
520 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
521 the first were themselves being wrongly included in the feed into dkim
522 processing; with most chunk sizes in use this resulted in an incorrect
523 body hash calculated value.
525 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
526 DKIM signature block, for verification. Although advised against by
527 standards it is specifically not ruled illegal.
529 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
531 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
532 missing a body hash (the bh= tag).
534 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
535 It seems that HAProxy sends the Proxy Protocol information in clear and
536 only then does a TLS startup, so do the same.
538 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
539 TCP connections (such as for Spamd) unless the daemon successfully set
540 Fast Open mode on its listening sockets. This fixes breakage seen on
541 too-old kernels or those not configured for Fast Open, at the cost of
542 requiring both directions being enabled for TFO, and TFO never being used
543 by non-daemon-related Exim processes.
545 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
546 endings, at least on the first header line. Try to canonify any that get
547 past that check, despite the cost.
549 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
550 now limited to an arbitrary five deep, while parsing addresses with the
551 strip_excess_angle_brackets option enabled.
553 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
554 instead leave the unprompted TLS handshake in socket buffer for the
555 TLS library to consume.
557 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
559 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
561 JH/16 Drop variables when they go out of scope. Memory management drops a whole
562 region in one operation, for speed, and this leaves assigned pointers
563 dangling. Add checks run only under the testsuite which checks all
564 variables at a store-reset and panics on a dangling pointer; add code
565 explicitly nulling out all the variables discovered. Fixes one known
566 bug: a transport crash, where a dangling pointer for $sending_ip_address
567 originally assigned in a verify callout, is re-used.
569 PP/06 Drop '.' from @INC in various Perl scripts.
571 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
573 PP/08 Reduce a number of compilation warnings under clang; building with
574 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
575 should be warning-free.
577 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
579 HS/01 Fix portability problems introduced by PP/08 for platforms where
580 realloc(NULL) is not equivalent to malloc() [SunOS et al].
582 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
583 chunk. This allows us to accept broken chunked messages. We need a more
584 general solution here.
586 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
587 already-broken messages in the queue.
589 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
591 JH/19 Fix reference counting bug in routing-generated-address tracking.
597 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
598 supports it and a size is available (ie. the sending peer gave us one).
600 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
601 years of being deprecated). The replacements are the ACLs
602 acl_smtp_mime and acl_not_smtp_mime.
604 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
605 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
606 or even an in-clear connection were permitted. Now, if the host lookup
607 was dnssec and dane was requested then the host is only used if the
608 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
609 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
610 if one fails this test.
611 This means that a poorly-configured remote DNS will make it incommunicado;
612 but it protects against a DNS-interception attack on it.
614 JH/04 Bug 1810: make continued-use of an open smtp transport connection
615 non-noisy when a race steals the message being considered.
617 JH/05 If main configuration option tls_certificate is unset, generate a
618 self-signed certificate for inbound TLS connections.
620 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
621 in rewrites and routers.
623 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
624 and logged a warning sing 4.83; now they are a configuration file error.
626 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
627 (lacking @domain). Apply the same qualification processing as RCPT.
629 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
631 JH/10 Support ${sha256:} applied to a string (as well as the previous
634 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
635 a cutthrough deliver is pending, as we always want to make a connection.
636 This also avoids re-routing the message when later placing the cutthrough
637 connection after a verify cache hit.
638 Do not update it with the verify result either.
640 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
641 when routing results in more than one destination address.
643 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
644 signing (which inhibits the cutthrough capability). Previously only
645 the presence of an option was tested; now an expansion evaluating as
646 empty is permissible (obviously it should depend only on data available
647 when the cutthrough connection is made).
649 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
650 the relevant preceding SMTP command did not note the pipelining mode.
652 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
653 Previously they were not counted.
655 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
656 as one having no matching records. Previously we deferred the message
657 that needed the lookup.
659 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
660 distinguished as "(=".
662 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
663 for missing MX records. Previously it only worked for missing A records.
665 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
667 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
668 after the data-go-ahead and data-ack. Patch from Jason Betts.
670 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
671 even for a "none" policy. Patch from Tony Meyer.
673 JH/22 Fix continued use of a connection for further deliveries. If a port was
674 specified by a router, it must also match for the delivery to be
677 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
678 When one of the recipients of a message was unsuitable for the connection
679 (has no matching addresses), we lost track of needing to mark it
680 deferred. As a result mail would be lost.
682 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
684 JH/25 Decoding ACL controls is now done using a binary search; the source code
685 takes up less space and should be simpler to maintain. Merge the ACL
686 condition decode tables also, with similar effect.
688 JH/26 Fix problem with one_time used on a redirect router which returned the
689 parent address unchanged. A retry would see the parent address marked as
690 delivered, so not attempt the (identical) child. As a result mail would
693 JH/27 Fix a possible security hole, wherein a process operating with the Exim
694 UID can gain a root shell. Credit to http://www.halfdog.net/ for
695 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
698 JH/28 Enable {spool,log} filesystem space and inode checks as default.
699 Main config options check_{log,spool}_{inodes,space} are now
700 100 inodes, 10MB unless set otherwise in the configuration.
702 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
703 Previously it only applied to the main-section connection policy
706 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
708 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
709 by me. Added RFC7919 DH primes as an alternative.
711 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
712 are not in the system include path.
714 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
715 GnuTLS, when a session startup failed (eg because the client disconnected)
716 Exim did stdio operations after fclose. This was exposed by a recent
717 change which nulled out the file handle after the fclose.
719 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
720 signed directly by the cert-signing cert, rather than an intermediate
721 OCSP-signing cert. This is the model used by LetsEncrypt.
723 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
725 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
726 an incoming connection.
728 HS/02 Bug 1802: Do not half-close the connection after sending a request
731 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
732 fallback to "prime256v1".
734 JH/34 SECURITY: Use proper copy of DATA command in error message.
735 Could leak key material. Remotely exploitable. CVE-2016-9963.
741 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
742 and 3.4.4 - once the server is enabled to respond to an OCSP request
743 it does even when not requested, resulting in a stapling non-aware
744 client dropping the TLS connection.
746 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
747 support variable-length bit vectors. No functional change.
749 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
750 The I= interface field on outgoing lines is now after the H= remote
751 host field, same as incoming lines. There is a separate
752 outgoing_interface log selector which allows you to disable the
755 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
756 If not running log_selector +smtp_connection the mainlog would be held
757 open indefinitely after a "too many connections" event, including to a
758 deleted file after a log rotate. Leave the per net connection logging
759 leaving it open for efficiency as that will be quickly detected by the
760 check on the next write.
762 HS/01 Bug 1671: Fix post transport crash.
763 Processing the wait-<transport> messages could crash the delivery
764 process if the message IDs didn't exist for some reason. When
765 using 'split_spool_directory=yes' the construction of the spool
766 file name failed already, exposing the same netto behaviour.
768 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
769 mime_regex ACL conditions.
771 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
772 to DSN fail messages (bounces): remote IP, remote greeting, remote response
773 to HELO, local diagnostic string.
775 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
776 log line to debug. Even when configured with a tls authenticator many
777 client connections are expected to not authenticate in this way, so
778 an authenticate fail is not an error.
780 HS/02 Add the Exim version string to the process info. This way exiwhat
781 gives some more detail about the running daemon.
783 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
784 matter for fast-change records such as DNSBLs.
786 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
787 retry record, even if constant. There may be multiple transports with
788 different interface settings and the retry behaviour needs to be kept
791 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
793 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
795 JH/10 Bug 840: fix log_defer_output option of pipe transport
797 JH/11 Bug 830: use same host for all RCPTS of a message, even under
798 hosts_randomize. This matters a lot when combined with mua_wrapper.
800 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
801 ${quote_pgsql:<string>} operator.
803 JH/13 Bug 1708: avoid misaligned access in cached lookup.
805 JH/14 Change header file name for freeradius-client. Relevant if compiling
806 with Radius support; from the Gentoo tree and checked under Fedora.
808 JH/15 Bug 1712: Introduce $prdr_requested flag variable
810 JH/16 Bug 1714: Permit an empty string as expansion result for transport
811 option transport_filter, meaning no filtering.
813 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
815 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
816 defaults to "*" (all hosts). The variable is now available when not built
817 with TLS, default unset, mainly to enable keeping the testsuite sane.
818 If a server certificate is not supplied (via tls_certificate) an error is
819 logged, and clients will find TLS connections fail on startup. Presumably
820 they will retry in-clear.
821 Packagers of Exim are strongly encouraged to create a server certificate
822 at installation time.
824 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
825 with the $config_file variable.
827 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
828 in transport context, after the attempt, and per-recipient. The latter type
829 is per host attempted. The event data is the error message, and the errno
830 information encodes the lookup type (A vs. MX) used for the (first) host,
831 and the trailing two digits of the smtp 4xx response.
833 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
834 to write to mainlog (or rejectlog, paniclog) in the window between file
835 creation and permissions/ownership being changed. Particularly affects
836 installations where exicyclog is run as root, rather than exim user;
837 result is that the running daemon panics and dies.
839 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
841 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
842 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
843 "pri" and "weight". Note that the previous implicit priority given by the
844 list order is no longer honoured.
846 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
849 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
850 by defining SUPPORT_SOCKS.
852 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
853 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
854 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
855 variables are renamed to proxy_{local,external}_{address,port}.
857 JH/27 Move Internationalisation support from Experimental to mainline, enabled
858 for a build by defining SUPPORT_I18N
860 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
861 of the query string, and make ${quote_redis:} do that quoting.
863 JH/29 Move Events support from Experimental to mainline, enabled by default
864 and removable for a build by defining DISABLE_EVENT.
866 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
868 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
869 cached by the daemon.
871 JH/32 Move Redis support from Experimental to mainline, enabled for a build
872 by defining LOOKUP_REDIS. The libhiredis library is required.
874 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
875 keys are given for lookup.
877 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
878 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
879 only supported when built with TLS support. The PolarSSL SHA routines
880 are still used when the TLS library is too old for convenient support.
882 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
883 openssl_options), for security. OpenSSL forces this from version 1.1.0
884 server-side so match that on older versions.
886 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
887 allocation for $value could be released as the expansion processing
888 concluded, but leaving the global pointer active for it.
890 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
891 and to use the domains and local_parts ACL conditions.
893 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
894 incorrectly not doubled on cutthrough transmission, hence seen as a
895 body-termination at the receiving system - resulting in truncated mails.
896 Commonly the sender saw a TCP-level error, and retransmitted the message
897 via the normal store-and-forward channel. This could result in duplicates
898 received - but deduplicating mailstores were liable to retain only the
899 initial truncated version.
901 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
903 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
905 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
906 we're in there, support oversigning also; bug 1309.
908 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
910 HS/04 Add support for keep_environment and add_environment options.
912 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
913 either intentional arithmetic overflow during PRNG, or testing config-
916 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
917 delivery resulted in actual delivery. Cancel cutthrough before DATA
920 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
921 rejects a recipient: pass the reject to the originator.
923 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
924 Many were false-positives and ignorable, but it's worth fixing the
927 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
928 for the new environment-manipulation done at startup. Move the routines
929 from being local to tls.c to being global via the os.c file.
931 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
932 an extract embedded as result-arg for a map, the first arg for extract
933 is unavailable so we cannot tell if this is a numbered or keyed
934 extraction. Accept either.
940 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
943 JH/02 The smtp transport option "multi_domain" is now expanded.
945 JH/03 The smtp transport now requests PRDR by default, if the server offers
948 JH/04 Certificate name checking on server certificates, when exim is a client,
949 is now done by default. The transport option tls_verify_cert_hostnames
950 can be used to disable this per-host. The build option
951 EXPERIMENTAL_CERTNAMES is withdrawn.
953 JH/05 The value of the tls_verify_certificates smtp transport and main options
954 default to the word "system" to access the system default CA bundle.
955 For GnuTLS, only version 3.0.20 or later.
957 JH/06 Verification of the server certificate for a TLS connection is now tried
958 (but not required) by default. The verification status is now logged by
959 default, for both outbound TLS and client-certificate supplying inbound
962 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
965 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
966 Status Notification (bounce) messages are now MIME format per RFC 3464.
967 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
968 under the control of the dsn_advertise_hosts option, and routers may
969 have a dsn_lasthop option.
971 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
972 default, modifiable by a malware= option. The list separator for
973 the options can now be changed in the usual way. Bug 68.
975 JH/10 The smtp_receive_timeout main option is now expanded before use.
977 JH/11 The incoming_interface log option now also enables logging of the
978 local interface on delivery outgoing connections.
980 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
981 if the interface and destination host and port all match.
983 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
986 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
987 Patch from Andrew Lewis.
989 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
990 now supports optional time-restrictions, weighting, and priority
991 modifiers per server. Patch originally by <rommer@active.by>.
993 JH/16 The spamd_address main option now supports a mixed list of local
994 and remote servers. Remote servers can be IPv6 addresses, and
995 specify a port-range.
997 JH/17 Bug 68: The spamd_address main option now supports an optional
998 timeout value per server.
1000 JH/18 Bug 1581: Router and transport options headers_add/remove can
1001 now have the list separator specified.
1003 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1006 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1009 JH/21 Support for the A6 type of dns record is withdrawn.
1011 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1012 rather than the verbs used.
1014 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1015 from 255 to 1024 chars.
1017 JH/24 Verification callouts now attempt to use TLS by default.
1019 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1020 are generic router options now. The defaults didn't change.
1022 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1023 Original patch from Alexander Shikoff, worked over by JH.
1025 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1026 files if ZCAT_COMMAND is not executable.
1028 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1030 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1032 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1033 Normally benign, it bites when the pair was led to by a CNAME;
1034 modern usage is to not canonicalize the domain to a CNAME target
1035 (and we were inconsistent anyway for A-only vs AAAA+A).
1037 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1039 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1040 when evaluating $sender_host_dnssec.
1042 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1043 $sender_helo_dnssec variable.
1045 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1047 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1049 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1051 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1052 documented as working, but never had. Support all but $spam_report.
1054 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1055 added for tls authenticator.
1057 HS/03 Add perl_taintmode main config option
1063 TL/01 When running the test suite, the README says that variables such as
1064 no_msglog_check are global and can be placed anywhere in a specific
1065 test's script, however it was observed that placement needed to be near
1066 the beginning for it to behave that way. Changed the runtest perl
1067 script to read through the entire script once to detect and set these
1068 variables, reset to the beginning of the script, and then run through
1069 the script parsing/test process like normal.
1071 TL/02 The BSD's have an arc4random API. One of the functions to induce
1072 adding randomness was arc4random_stir(), but it has been removed in
1073 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1074 function when detected.
1076 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1077 cause callback expansion.
1079 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1080 syntax errors in an expansion can be treated as a string instead of
1081 logging or causing an error, due to the internal use of bool_lax
1082 instead of bool when processing it.
1084 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1085 server certificates when making smtp deliveries.
1087 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1089 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1091 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1093 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1094 Merged patch from Sebastian Wiedenroth.
1096 JH/05 Fix results-pipe from transport process. Several recipients, combined
1097 with certificate use, exposed issues where response data items split
1098 over buffer boundaries were not parsed properly. This eventually
1099 resulted in duplicates being sent. This issue only became common enough
1100 to notice due to the introduction of connection certificate information,
1101 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1103 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1104 size buffer was used, resulting in syntax errors when an expansion
1107 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1108 version 3.3.6 or later.
1110 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1111 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1112 both become "event_action", the variables become $event_name, $event_data
1113 and $event_defer_errno. There is a new variable $verify_mode, usable in
1114 routers, transports and related events. The tls:cert event is now also
1115 raised for inbound connections, if the main configuration event_action
1118 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1119 early OCSP support, but no stapling (appears to be less than 1.0.0).
1121 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1122 server certificate names available under the smtp transport option
1123 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1126 JH/10 Time-related extraction expansions from certificates now use the main
1127 option "timezone" setting for output formatting, and are consistent
1128 between OpenSSL and GnuTLS compilations. Bug 1541.
1130 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1131 encoded parameter in the incoming message. Bug 1558.
1133 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1134 include certificate info, eximon was claiming there were spoolfile
1137 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1139 JH/14 Log delivery-related information more consistently, using the sequence
1140 "H=<name> [<ip>]" wherever possible.
1142 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1143 are problematic for Debian distribution, omit them from the release
1146 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1148 JH/16 Fix string representation of time values on 64bit time_t architectures.
1151 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1152 output list separator was used.
1157 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1158 checkers that were complaining about end of non-void function with no
1161 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1162 This was a regression introduced in 4.83 by another bugfix.
1164 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1166 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1167 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1173 TF/01 Correctly close the server side of TLS when forking for delivery.
1175 When a message was received over SMTP with TLS, Exim failed to clear up
1176 the incoming connection properly after forking off the child process to
1177 deliver the message. In some situations the subsequent outgoing
1178 delivery connection happened to have the same fd number as the incoming
1179 connection previously had. Exim would try to use TLS and fail, logging
1180 a "Bad file descriptor" error.
1182 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1183 utilities have not been installed.
1185 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1186 temporary space as the ACL may create new global variables.
1188 TL/01 LDAP support uses per connection or global context settings, depending
1189 upon the detected version of the libraries at build time.
1191 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1192 to extract and use the src ip:port in logging and expansions as if it
1193 were a direct connection from the outside internet. PPv2 support was
1194 updated based on HAProxy spec change in May 2014.
1196 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1198 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1199 Properly escape header and check for NULL return.
1201 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1204 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1206 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1207 characters in header names, implemented as a verify condition.
1208 Contributed by Michael Fischer v. Mollard.
1210 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1211 results permerror and temperror. Previous values are deprecated but
1212 still accepted. In a future release, err_perm and err_temp will be
1213 completely removed, which will be a backward incompatibility if the
1214 ACL tests for either of these two old results. Patch contributed by
1215 user bes-internal on the mailing list.
1217 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1219 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1220 selectors, in both main and reject logs.
1222 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1225 JH/07 Add malware type "sock" for talking to simple daemon.
1227 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1229 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1230 routers/transports under cutthrough routing.
1232 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1233 numbers. Touch up "bool" conditional to keep the same definition.
1235 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1237 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1239 JH/12 Expand items in router/transport headers_add or headers_remove lists
1240 individually rather than the list as a whole. Bug 1452.
1242 Required for reasonable handling of multiple headers_ options when
1243 they may be empty; requires that headers_remove items with embedded
1244 colons must have them doubled (or the list-separator changed).
1246 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1247 view the policy declared in the DMARC record. Currently, $dmarc_status
1248 is a combined value of both the record presence and the result of the
1251 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1253 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1254 dnslookup router and the smtp transport (applying to the forward
1257 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1258 of ldap servers used for a specific lookup. Patch provided by Heiko
1261 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1262 New variable $lookup_dnssec_authenticated for observability.
1264 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1265 Patch submitted by Lars Timman.
1267 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1269 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1270 Requires trusted mode and valid format message id, aborts otherwise.
1271 Patch contributed by Heiko Schlichting.
1273 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1274 certextract with support for various fields. Bug 1358.
1276 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1277 is requested by default, modifiable by smtp transport option
1280 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1281 operate on certificate variables to give certificate fingerprints
1282 Also new ${sha256:cert_variable}.
1284 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1286 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1287 Christian Aistleitner.
1289 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1291 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1292 file. Patch from Wolfgang Breyha.
1294 JH/25 Expand the coverage of the delivery $host and $host_address to
1295 client authenticators run in verify callout. Bug 1476.
1297 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1298 align with daemon_smtp_ports. Bug 72.
1300 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1301 support and error reporting did not work properly.
1303 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1304 and is readable. Patch from Andrew Colin Kissa.
1306 TL/14 Enhance documentation of ${run expansion and how it parses the
1307 commandline after expansion, particularly in the case when an
1308 unquoted variable expansion results in an empty value.
1310 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1312 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1313 by renaming away from it. Addresses GH issue 15, reported by
1316 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1318 TL/15 SECURITY: prevent double expansion in math comparison functions
1319 (can expand unsanitized data). Not remotely exploitable.
1326 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1328 PP/02 Make -n do something, by making it not do something.
1329 When combined with -bP, the name of an option is not output.
1331 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1334 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1335 $sender_host_name and config options to manage this, and basic check
1338 PP/05 DSCP support for outbound connections and control modifier for inbound.
1340 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1341 (Only plugin which currently uses this is kerberos4, which nobody should
1342 be using, but we should make it available and other future plugins might
1343 conceivably use it, even though it would break NAT; stuff *should* be
1344 using channel bindings instead).
1346 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1347 name; added for Sendmail compatibility; requires admin caller.
1348 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1349 just ignore it); requires trusted caller.
1350 Also parse but ignore: -Ac -Am -X<logfile>
1353 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1355 TL/02 Add +smtp_confirmation as a default logging option.
1357 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1358 Patch by Magnus Holmgren from 2007-02-20.
1360 TL/04 Bugzilla 1281 - Spec typo.
1361 Bugzilla 1283 - Spec typo.
1362 Bugzilla 1290 - Spec grammar fixes.
1364 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1366 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1368 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1369 mailing list by by Dmitry Isaikin.
1371 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1373 JH/02 Support "G" suffix to numbers in ${if comparisons.
1375 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1377 NM/01 Bugzilla 1197 - Spec typo
1378 Bugzilla 1196 - Spec examples corrections
1380 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1382 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1383 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1386 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1387 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1389 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1390 "acl {{name}{arg}...}", and optional args on acl condition
1393 JH/05 Permit multiple router/transport headers_add/remove lines.
1395 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1397 JH/07 Avoid using a waiting database for a single-message-only transport.
1398 Performance patch from Paul Fisher. Bugzilla 1262.
1400 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1403 JH/09 Add $headers_added variable, with content from use of ACL modifier
1404 add_header (but not yet added to the message). Bugzilla 199.
1406 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1407 Pulled from Bugzilla 817 by Wolfgang Breyha.
1409 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1411 (nb: this is the same fix as in Exim 4.80.1)
1413 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1416 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1419 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1421 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1422 advertise SMTP AUTH mechanism to us, instead of a generic
1423 protocol violation error. Also, make Exim more robust to bad
1424 data from the Dovecot auth socket.
1426 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1428 When a queue runner is handling a message, Exim first routes the
1429 recipient addresses, during which it prunes them based on the retry
1430 hints database. After that it attempts to deliver the message to
1431 any remaining recipients. It then updates the hints database using
1434 So if a recipient address works intermittently, it can get repeatedly
1435 deferred at routing time. The retry hints record remains fresh so the
1436 address never reaches the final cutoff time.
1438 This is a fairly common occurrence when a user is bumping up against
1439 their storage quota. Exim had some logic in its local delivery code
1440 to deal with this. However it did not apply to per-recipient defers
1441 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1443 This change adds a proper retry rule check during routing so that the
1444 final cutoff time is checked against the message's age. We only do
1445 this check if there is an address retry record and there is not a
1446 domain retry record; this implies that previous attempts to handle
1447 the address had the retry_use_local_parts option turned on. We use
1448 this as an approximation for the destination being like a local
1449 delivery, as in LMTP.
1451 I suspect this new check makes the old local delivery cutoff check
1452 redundant, but I have not verified this so I left the code in place.
1454 TF/02 Correct gecos expansion when From: is a prefix of the username.
1456 Test 0254 submits a message to Exim with the header
1460 When I ran the test suite under the user fanf2, Exim expanded
1461 the header to contain my full name, whereas it should have added
1462 a Resent-Sender: header. It erroneously treats any prefix of the
1463 username as equal to the username.
1465 This change corrects that bug.
1467 GF/01 DCC debug and logging tidyup
1468 Error conditions log to paniclog rather than rejectlog.
1469 Debug lines prefixed by "DCC: " to remove any ambiguity.
1471 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1473 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1474 Bug spotted by Jeremy Harris; was flawed since initial commit.
1475 Would have resulted in OCSP responses post-SNI triggering an Exim
1476 NULL dereference and crash.
1478 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1480 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1481 Bug detection, analysis and fix by Samuel Thibault.
1482 Bugzilla 1331, Debian bug #698092.
1484 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1486 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1487 Server implementation by Todd Lyons, client by JH.
1488 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1489 config variable "prdr_enable" controls whether the server
1490 advertises the facility. If the client requests PRDR a new
1491 acl_data_smtp_prdr ACL is called once for each recipient, after
1492 the body content is received and before the acl_smtp_data ACL.
1493 The client is controlled by both of: a hosts_try_prdr option
1494 on the smtp transport, and the server advertisement.
1495 Default client logging of deliveries and rejections involving
1496 PRDR are flagged with the string "PRDR".
1498 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1499 fclose(). Diagnosis by Todd Lyons.
1501 PP/17 Update configure.default to handle IPv6 localhost better.
1502 Patch by Alain Williams (plus minor tweaks).
1505 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1506 This is now consistent with GnuTLS, and is now documented: the
1507 previous undocumented portable approach to treating the option as
1508 unset was to force an expansion failure. That still works, and
1509 an empty string is now equivalent.
1511 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1512 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1513 not performing validation itself.
1515 PP/20 Added force_command boolean option to pipe transport.
1516 Patch from Nick Koston, of cPanel Inc.
1518 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1521 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1523 PP/21 Fix eximon continuous updating with timestamped log-files.
1524 Broken in a format-string cleanup in 4.80, missed when I repaired the
1525 other false fix of the same issue.
1526 Report and fix from Heiko Schlichting.
1529 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1530 Report from Prashanth Katuri.
1532 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1533 It's SecureTransport, so affects any MacOS clients which use the
1534 system-integrated TLS libraries, including email clients.
1536 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1537 using a MIME ACL for non-SMTP local injection.
1538 Report and assistance in diagnosis by Warren Baker.
1540 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1542 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1544 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1545 last id that failed so it may be referenced in subsequent ACL's.
1547 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1550 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1551 ldap library initialization, allowing self-signed CA's to be
1552 used. Also properly sets require_cert option later in code by
1553 using NULL (global ldap config) instead of ldap handle (per
1554 session). Bug diagnosis and testing by alxgomz.
1556 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1557 the src/util/ subdirectory.
1559 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1560 renamed to Transport Post Delivery Action by Jeremy Harris, as
1563 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1564 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1565 redis_servers = needs to be configured which will be used by the redis
1566 lookup. Patch from Warren Baker, of The Packet Hub.
1568 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1570 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1571 hostname or reverse DNS when processing a host list. Used suggestions
1572 from multiple comments on this bug.
1574 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1576 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1577 Missed a few lines, added it to make the runtest require no keyboard
1580 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1581 contains upper case chars. Make router use caseful_local_part.
1583 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1584 support when GnuTLS has been built with p11-kit.
1590 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1592 This, or similar/improved, will also be change PP/11 of 4.82.
1598 PP/01 Handle short writes when writing local log-files.
1599 In practice, only affects FreeBSD (8 onwards).
1600 Bugzilla 1053, with thanks to Dmitry Isaikin.
1602 NM/01 Bugzilla 949 - Documentation tweak
1604 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1607 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1609 PP/02 Implemented gsasl authenticator.
1611 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1613 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1614 `pkg-config foo` for cflags/libs.
1616 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1617 with rest of GSASL and with heimdal_gssapi.
1619 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1620 `pkg-config foo` for cflags/libs for the TLS implementation.
1622 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1623 properties get this fed in as external SSF. A number of robustness
1624 and debugging improvements to the cyrus_sasl authenticator.
1626 PP/08 cyrus_sasl server now expands the server_realm option.
1628 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1629 Patch by Jeremy Harris.
1631 PP/10 Added dbmjz lookup type.
1633 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1635 PP/12 MAIL args handles TAB as well as SP, for better interop with
1636 non-compliant senders.
1637 Analysis and variant patch by Todd Lyons.
1639 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1640 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1641 Patch from Dirk Mueller <dmueller@suse.com>
1643 PP/13 tls_peerdn now print-escaped for spool files.
1644 Observed some $tls_peerdn in wild which contained \n, which resulted
1645 in spool file corruption.
1647 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1648 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1649 or write after TLS renegotiation, which otherwise led to messages
1652 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1653 as a tracking header (ie: a signed header comes before the signature).
1654 Patch from Wolfgang Breyha.
1656 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1657 comma-sep list; embedded commas doubled.
1659 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1661 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1663 Report and patch from Dmitry Banschikov.
1665 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1666 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1667 failures. We appear to now support TLS1.1+ with Exim.
1669 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1670 lets Exim select keys and certificates based upon TLS SNI from client.
1671 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1672 before an outbound SMTP session. New log_selector, +tls_sni.
1674 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1675 NULL dereference. Report and patch from Alun Jones.
1677 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1678 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1679 Not seeing resolver debug output on NetBSD, but suspect this is a
1680 resolver implementation change.
1682 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1683 Left warnings. Added "eximon gdb" invocation mode.
1685 PP/21 Defaulting "accept_8bitmime" to true, not false.
1687 PP/22 Added -bw for inetd wait mode support.
1689 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1690 locate the relevant includes and libraries. Made this the default.
1692 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1693 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1695 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1696 This may cause build issues on older platforms.
1698 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1699 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1700 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1701 Added SNI support via GnuTLS too.
1702 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1704 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1706 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1707 Applied second patch from Janne, implementing suggestion to default
1708 multiple-strings-in-record handling to match SPF spec.
1710 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1712 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1713 read-only, out of scope).
1714 Patch from Wolfgang Breyha, report from Stuart Northfield.
1716 PP/29 Fix three issues highlighted by clang analyser static analysis.
1717 Only crash-plausible issue would require the Cambridge-specific
1718 iplookup router and a misconfiguration.
1719 Report from Marcin Mirosław.
1721 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1723 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1724 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1725 As part of this, removing so much warning spew let me fix some minor
1726 real issues in debug logging.
1728 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1729 assignment on my part. Fixed.
1731 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1732 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1733 Janne Snabb (who went above and beyond: thank you).
1735 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1736 string otherwise requires a connection and a bunch more work and it's
1737 relatively easy to get wrong. Should also expose TLS library linkage
1740 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1741 64-bit ${eval} (JH/03).
1743 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1744 GNU libc to support some of the 64-bit stuff, should not lead to
1745 conflicts. Defined before os.h is pulled in, so if a given platform
1746 needs to override this, it can.
1748 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1749 protection layer was required, which is not implemented.
1750 Bugzilla 1254, patch from Wolfgang Breyha.
1752 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1753 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1754 tls_dhparam take prime identifiers. Also unbreak combination of
1755 OpenSSL+DH_params+TLSSNI.
1757 PP/39 Disable SSLv2 by default in OpenSSL support.
1763 PP/01 Solaris build fix for Oracle's LDAP libraries.
1764 Bugzilla 1109, patch from Stephen Usher.
1766 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1768 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1771 TF/02 Fix a couple more cases where we did not log the error message
1772 when unlink() failed. See also change 4.74-TF/03.
1774 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1775 lock up or crash if it happened to be inside a call to libc when it
1776 got a SIGUSR1 from exiwhat.
1778 The SIGUSR1 handler appends the current process status to the process
1779 log which is later printed by exiwhat. It used to use the general
1780 purpose logging code to do this, but several functions it calls are
1781 not safe for signals.
1783 The new output code in the SIGUSR1 handler is specific to the process
1784 log, and simple enough that it's easy to inspect for signal safety.
1785 Removing some special cases also simplifies the general logging code.
1786 Removing the spurious timestamps from the process log simplifies
1789 TF/04 Improved ratelimit ACL condition.
1791 The /noupdate option has been deprecated in favour of /readonly which
1792 has clearer semantics. The /leaky, /strict, and /readonly update modes
1793 are mutually exclusive. The update mode is no longer included in the
1794 database key; it just determines when the database is updated. (This
1795 means that when you upgrade Exim will forget old rate measurements.)
1797 Exim now checks that the per_* options are used with an update mode that
1798 makes sense for the current ACL. For example, when Exim is processing a
1799 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1800 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1801 must specify per_mail/readonly. If you omit the update mode it defaults to
1802 /leaky where that makes sense (as before) or /readonly where required.
1804 The /noupdate option is now undocumented but still supported for
1805 backwards compatibility. It is equivalent to /readonly except that in
1806 ACLs where /readonly is required you may specify /leaky/noupdate or
1807 /strict/noupdate which are treated the same as /readonly.
1809 A useful new feature is the /count= option. This is a generalization
1810 of the per_byte option, so that you can measure the throughput of other
1811 aggregate values. For example, the per_byte option is now equivalent
1812 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1814 The per_rcpt option has been generalized using the /count= mechanism
1815 (though it's more complicated than the per_byte equivalence). When it is
1816 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1817 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1818 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1819 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1820 Note that using per_rcpt with a non-readonly update mode in more than
1821 one ACL will cause the recipients to be double-counted. (The per_mail
1822 and per_byte options don't have this problem.)
1824 The handling of very low rates has changed slightly. If the computed rate
1825 is less than the event's count (usually one) then this event is the first
1826 after a long gap. In this case the rate is set to the same as this event's
1827 count, so that the first message of a spam run is counted properly.
1829 The major new feature is a mechanism for counting the rate of unique
1830 events. The new per_addr option counts the number of different
1831 recipients that someone has sent messages to in the last time period. It
1832 behaves like per_rcpt if all the recipient addresses are different, but
1833 duplicate recipient addresses do not increase the measured rate. Like
1834 the /count= option this is a general mechanism, so the per_addr option
1835 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1836 example, measure the rate that a client uses different sender addresses
1837 with the options per_mail/unique=$sender_address. There are further
1838 details in the main documentation.
1840 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1842 TF/06 Removed a few PCRE remnants.
1844 TF/07 Automatically extract Exim's version number from tags in the git
1845 repository when doing development or release builds.
1847 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1848 Bugzilla 879. Patch from Paul Fisher.
1850 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1851 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1854 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1856 PP/05 Variable $av_failed, true if the AV scanner deferred.
1857 Bugzilla 1078. Patch from John Horne.
1859 PP/06 Stop make process more reliably on build failure.
1860 Bugzilla 1087. Patch from Heiko Schlittermann.
1862 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1863 Bugzilla 1089. Patch from Heiko Schlittermann.
1865 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1866 Bugzilla 1131. Patch from Holger Weiß.
1868 PP/09 Handle IPv6 addresses with SPF.
1869 Bugzilla 860. Patch from Wolfgang Breyha.
1871 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1873 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1876 PP/11 match_* no longer expand right-hand-side by default.
1877 New compile-time build option, EXPAND_LISTMATCH_RHS.
1878 New expansion conditions, "inlist", "inlisti".
1880 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1882 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1884 PP/14 fix log_write() format string regression from TF/03.
1885 Bugzilla 1152. Patch from Dmitry Isaikin.
1891 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1893 PP/02 Harmonised TLS library version reporting; only show if debugging.
1894 Layout now matches that introduced for other libraries in 4.74 PP/03.
1896 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1898 PP/04 New "dns_use_edns0" global option.
1900 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1903 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1904 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1906 TK/01 Updated PolarSSL code to 0.14.2.
1907 Bugzilla 1097. Patch from Andreas Metzler.
1909 PP/07 Catch divide-by-zero in ${eval:...}.
1910 Fixes bugzilla 1102.
1912 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1915 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1916 format-string attack -- SECURITY: remote arbitrary code execution.
1918 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1919 time unintentionally subject to list matching rules, letting the header
1920 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1921 arbitrary string expansion). This allowed for information disclosure.
1923 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1924 INT_MIN/-1 -- value coerced to INT_MAX.
1930 NM/01 Workaround for PCRE version dependency in version reporting
1933 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1934 This fixes portability to compilers other than gcc, notably
1935 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1937 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1938 makefiles for portability to HP-UX and POSIX correctness.
1940 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1941 Also via indented variable definition in the Makefile.
1942 (Debugging by Oliver Heesakkers).
1944 PP/02 Restore caching of spamd results with expanded spamd_address.
1945 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1947 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1948 Improves build reliability. Fix from: Frank Elsner
1950 NM/02 Fix wide character breakage in the rfc2047 coding
1951 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1953 NM/03 Allow underscore in dnslist lookups
1954 Fixes bug 1026. Patch from Graeme Fowler
1956 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1957 Code patches from Adam Ciarcinski of NetBSD.
1959 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1962 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1963 is logged, to avoid truncation. Patch from John Horne.
1965 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1966 Patch from Jakob Hirsch.
1968 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1969 SQL string expansion failure details.
1970 Patch from Andrey Oktyabrski.
1972 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1973 Patch from Simon Arlott.
1975 PP/09 New lookups functionality failed to compile on old gcc which rejects
1976 extern declarations in function scope.
1977 Patch from Oliver Fleischmann
1979 PP/10 Use sig_atomic_t for flags set from signal handlers.
1980 Check getgroups() return and improve debugging.
1981 Fixed developed for diagnosis in bug 927 (which turned out to be
1984 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1985 Patch from Mark Zealey.
1987 PP/12 Bugzilla 1056: Improved spamd server selection.
1988 Patch from Mark Zealey.
1990 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1991 Based on patch from Heiko Schlittermann.
1993 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1994 Patch from Uwe Doering, sign-off by Michael Haardt.
1996 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1997 variable declaration deep within a block. Bug and patch from
2000 PP/15 lookups-Makefile IRIX compatibility coercion.
2002 PP/16 Make DISABLE_DKIM build knob functional.
2004 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2005 Patch by Simon Arlott
2007 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2008 variable argument macros. Our copy now differs from upstream.
2014 TF/01 Failure to get a lock on a hints database can have serious
2015 consequences so log it to the panic log.
2017 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2018 controlled using the smtp_confirmation log selector.
2020 TF/03 Include the error message when we fail to unlink a spool file.
2022 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2023 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2024 for maintaining out-of-tree patches for some time.
2026 PP/01 Bugzilla 139: Documentation and portability issues.
2027 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2028 Handle per-OS dynamic-module compilation flags.
2030 PP/02 Let /dev/null have normal permissions.
2031 The 4.73 fixes were a little too stringent and complained about the
2032 permissions on /dev/null. Exempt it from some checks.
2033 Reported by Andreas M. Kirchwitz.
2035 PP/03 Report version information for many libraries, including
2036 Exim version information for dynamically loaded libraries. Created
2037 version.h, now support a version extension string for distributors
2038 who patch heavily. Dynamic module ABI change.
2040 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2041 privilege escalation vulnerability whereby the Exim run-time user
2042 can cause root to append content of the attacker's choosing to
2045 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2048 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2049 If dropping privileges for untrusted macros, we disabled normal logging
2050 on the basis that it would fail; for the Exim run-time user, this is not
2051 the case, and it resulted in successful deliveries going unlogged.
2052 Fixed. Reported by Andreas Metzler.
2058 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2059 only prepend for the Resent-* case. Fixes regression introduced in
2060 Exim 4.70 by NM/22 for Bugzilla 607.
2062 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2063 increasing numbers of administrators be bitten by this.
2065 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2067 PP/03 Bugzilla 994: added openssl_options main configuration option.
2069 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2071 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2073 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2075 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2076 without a peer certificate, leading to a segfault because of an
2077 assumption that peers always have certificates. Be a little more
2078 paranoid. Problem reported by Martin Tscholak.
2080 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2081 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2082 NB: ClamAV planning to remove STREAM in "middle of 2010".
2083 CL also introduces -bmalware, various -d+acl logging additions and
2084 more caution in buffer sizes.
2086 PP/09 Implemented reverse_ip expansion operator.
2088 PP/10 Bugzilla 937: provide a "debug" ACL control.
2090 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2092 PP/12 Bugzilla 973: Implement --version.
2094 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2096 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2098 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2100 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2101 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2102 ignore trailing whitespace.
2104 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2106 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2109 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2110 Notification from Dr Andrew Aitchison.
2112 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2113 ExtendedDetectionInfo response format.
2114 Notification from John Horne.
2116 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2119 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2120 XSL and documented dependency on system catalogs, with examples of how
2123 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2126 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2127 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2128 configuration file which is writeable by the Exim user or group.
2130 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2131 of configuration files to cover files specified with the -C option if
2132 they are going to be used with root privileges, not just the default
2135 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2136 option (effectively making it always true).
2138 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2139 files to be used while preserving root privileges.
2141 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2142 that rogue child processes cannot use them.
2144 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2145 run-time user, instead of root.
2147 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2148 Exim run-time user without dropping privileges.
2150 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2151 result string, instead of calling string_vformat() twice with the same
2154 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2155 for other users. Others should always drop root privileges if they use
2156 -C on the command line, even for a whitelisted configure file.
2158 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2160 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2166 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2167 $data_path, and $header_path variables; fixed documentation bugs and
2170 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2171 exipick to access non-standard spools, including the "frozen" queue
2174 NM/01 Bugzilla 965: Support mysql stored procedures.
2175 Patch from Alain Williams
2177 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2179 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2180 Patch from Andreas Metzler
2182 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2183 Patch from Kirill Miazine
2185 NM/05 Bugzilla 671: Added umask to procmail example.
2187 JJ/03 installed exipick 20100323.0, fixing doc bug
2189 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2190 directory. Notification and patch from Dan Rosenberg.
2192 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2194 TK/02 Improve log output when DKIM signing operation fails.
2196 MH/01 Treat the transport option dkim_domain as a colon separated
2197 list, not as a single string, and sign the message with each element,
2198 omitting multiple occurences of the same signer.
2200 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2201 Bugzilla 985, 986. Patch by Simon Arlott
2203 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2204 Patch by Simon Arlott
2206 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2207 MBX locking. Notification from Dan Rosenberg.
2213 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2215 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2217 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2219 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2221 NM/04 Bugzilla 915: Flag broken perl installation during build.
2227 TK/01 Added patch by Johannes Berg that expands the main option
2228 "spamd_address" if it starts with a dollar sign.
2230 TK/02 Write list of recipients to X-Envelope-Sender header when building
2231 the mbox-format spool file for content scanning (suggested by Jakob
2234 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2235 (http://www.dcc-servers.net/) support via dccifd. Activated by
2236 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2238 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2239 by Mark Daniel Reidel <mr@df.eu>.
2241 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2242 When building exim an external PCRE library is now needed -
2243 PCRE is a system library on the majority of modern systems.
2244 See entry on PCRE_LIBS in EDITME file.
2246 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2247 conversation. Added nologin parameter to request.
2248 Patch contributed by Kirill Miazine.
2250 TF/01 Do not log submission mode rewrites if they do not change the address.
2252 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2254 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2255 log files in place. Contributed by Roberto Lima.
2257 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2259 TF/03 Bugzilla 615: When checking the local_parts router precondition
2260 after a local_part_suffix or local_part_prefix option, Exim now
2261 does not use the address's named list lookup cache, since this
2262 contains cached lookups for the whole local part.
2264 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2265 Robert Millan. Documentation is in experimental-spec.txt.
2267 TF/04 Bugzilla 668: Fix parallel build (make -j).
2269 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2271 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2272 Patch provided by Jan Srzednicki.
2274 TF/05 Leading white space used to be stripped from $spam_report which
2275 wrecked the formatting. Now it is preserved.
2277 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2278 that they are available at delivery time.
2280 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2282 TF/08 TLS error reporting now respects the incoming_interface and
2283 incoming_port log selectors.
2285 TF/09 Produce a more useful error message if an SMTP transport's hosts
2286 setting expands to an empty string.
2288 NM/06 Bugzilla 744: EXPN did not work under TLS.
2289 Patch provided by Phil Pennock.
2291 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2292 Patch provided by Richard Godbee.
2294 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2295 acl_smtp_notquit, added index entry.
2297 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2298 Patch provided by Eugene Bujak.
2300 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2301 accept(). Patch provided by Maxim Dounin.
2303 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2304 Patch provided by Phil Pennock.
2306 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2308 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2309 Patch provided by Brad "anomie" Jorsch.
2311 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2312 Patch provided by Dean Brooks.
2314 TK/05 Add native DKIM support (does not depend on external libraries).
2316 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2317 Patch provided by Graeme Fowler.
2319 NM/16 Bugzilla 851: Documentation example syntax fix.
2321 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2323 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2326 NM/19 Bugzilla 745: TLS version reporting.
2327 Patch provided by Phil Pennock.
2329 NM/20 Bugzilla 167: bool: condition support.
2330 Patch provided by Phil Pennock.
2332 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2333 clients. Patch provided by Phil Pennock.
2335 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2336 Patch provided by Brad "anomie" Jorsch.
2338 NM/23 Bugzilla 687: Fix misparses in eximstats.
2339 Patch provided by Heiko Schlittermann.
2341 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2342 Patch provided by Heiko Schlittermann.
2344 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2345 plus update to original patch.
2347 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2349 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2350 Patch provided by David Brownlee.
2352 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2354 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2356 NM/30 Bugzilla 888: TLS documentation bugfixes.
2358 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2360 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2361 Unlike the original bugzilla I have changed all shell scripts in src tree.
2363 NM/33 Bugzilla 898: Transport filter timeout fix.
2364 Patch by Todd Rinaldo.
2366 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2367 Patch by Serge Demonchaux.
2369 NM/35 Bugzilla 39: Base64 decode bug fixes.
2370 Patch by Jakob Hirsch.
2372 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2374 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2376 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2378 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2384 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2385 ALT-N's libdkim that I have put here:
2386 http://duncanthrax.net/exim-experimental/
2388 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2389 were called 'true' and it seems that C99 defines that as a reserved
2390 keyword to be used with 'bool' variable types. That means you could
2391 not include C99-style headers which use bools without triggering
2392 build errors in sieve.c.
2394 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2395 as mailq or other aliases. Changed the --help handling significantly
2396 to do whats expected. exim_usage() emits usage/help information.
2398 SC/01 Added the -bylocaldomain option to eximstats.
2400 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2402 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2404 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2410 PH/01 Another patch from the Sieve maintainer.
2412 PH/02 When an IPv6 address is converted to a string for single-key lookup
2413 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2414 dots are used instead of colons so that keys in lsearch files need not
2415 contain colons. This was done some time before quoting was made available
2416 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2417 (notated using the quote facility) so as to distinguish them from IPv4
2418 keys. This meant that lookups for IP addresses in host lists did not work
2419 for iplsearch lookups.
2421 This has been fixed by arranging for IPv6 addresses to be expressed with
2422 colons if the lookup type is iplsearch. This is not incompatible, because
2423 previously such lookups could never work.
2425 The situation is now rather anomalous, since one *can* have colons in
2426 ordinary lsearch keys. However, making the change in all cases is
2427 incompatible and would probably break a number of configurations.
2429 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2432 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2433 conversion specification without a maximum field width, thereby enabling
2434 a rogue spamd server to cause a buffer overflow. While nobody in their
2435 right mind would setup Exim to query an untrusted spamd server, an
2436 attacker that gains access to a server running spamd could potentially
2437 exploit this vulnerability to run arbitrary code as the Exim user.
2439 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2440 $primary_hostname instead of what libspf2 thinks the hosts name is.
2442 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2443 a directory entry by the name of the lookup key. Previously, if a
2444 symlink pointed to a non-existing file or a file in a directory that
2445 Exim lacked permissions to read, a lookup for a key matching that
2446 symlink would fail. Now it is enough that a matching directory entry
2447 exists, symlink or not. (Bugzilla 503.)
2449 PH/03 The body_linecount and body_zerocount variables are now exported in the
2452 PH/04 Added the $dnslist_matched variable.
2454 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2455 This means they are set thereafter only if the connection becomes
2458 PH/06 Added the client_condition to authenticators so that some can be skipped
2459 by clients under certain conditions.
2461 PH/07 The error message for a badly-placed control=no_multiline_responses left
2462 "_responses" off the end of the name.
2464 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2466 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2467 (without spaces) instead of just copying the configuration text.
2469 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2471 PH/11 Added $max_received_linelength.
2473 PH/12 Added +ignore_defer and +include_defer to host lists.
2475 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2476 way in which PCRE > 7.0 is built.
2478 PH/14 Implemented queue_only_load_latch.
2480 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2481 MAIL command. The effect was to mangle the value on 64-bit systems.
2483 PH/16 Another patch from the Sieve maintainer.
2485 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2487 PH/18 If a system quota error occurred while trying to create the file for
2488 a maildir delivery, the message "Mailbox is full" was not appended to the
2489 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2490 only to a quota excession during the actual writing of the file.
2492 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2493 characters?) which causes problems in log lines. The DN values are now
2494 passed through string_printing() before being added to log lines.
2496 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2497 and InterBase are left for another time.)
2499 PH/21 Added message_body_newlines option.
2501 PH/22 Guard against possible overflow in moan_check_errorcopy().
2503 PH/23 POSIX allows open() to be a macro; guard against that.
2505 PH/24 If the recipient of an error message contained an @ in the local part
2506 (suitably quoted, of course), incorrect values were put in $domain and
2507 $local_part during the evaluation of errors_copy.
2513 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2514 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2517 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2518 issue a MAIL command.
2520 PH/02 In an ACL statement such as
2522 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2524 if a client was not listed at all, or was listed with a value other than
2525 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2526 the condition was not true (as it should be), so access was not denied.
2527 The bug was that the ! inversion was incorrectly passed on to the second
2528 item. This has been fixed.
2530 PH/03 Added additional dnslists conditions == and =& which are different from
2531 = and & when the dns lookup returns more than one IP address.
2533 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2534 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2536 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2537 FSYNC, which compiles an option called disable_fsync that allows for
2538 bypassing fsync(). The documentation is heavily laced with warnings.
2540 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2542 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2543 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2544 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2545 including adding "make clean"; (3) Added -fPIC when compiling the test
2546 dynamically loaded module, to get rid of a warning.
2548 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2549 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2550 = 0s. The bug is otherwise harmless.
2552 PH/07 There was a bug in the dovecot authenticator such that the value of
2553 $auth1 could be overwritten, and so not correctly preserved, after a
2554 successful authentication. This usually meant that the value preserved by
2555 the server_setid option was incorrect.
2557 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2559 PH/09 Installed PCRE release 7.0.
2561 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2562 run for batched SMTP input. It is now run at the start of every message
2563 in the batch. While fixing this I discovered that the process information
2564 (output by running exiwhat) was not always getting set for -bs and -bS
2565 input. This is fixed, and it now also says "batched" for BSMTP.
2567 PH/11 Added control=no_pipelining.
2569 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2570 patch, slightly modified), and move the expansion of helo_data till after
2571 the connection is made in the smtp transport (so it can use these
2574 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2576 PH/14 Added log_selector = +pid.
2578 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2580 PH/16 Add ${if forany and ${if forall.
2582 PH/17 Added dsn_from option to vary the From: line in DSNs.
2584 PH/18 Flush SMTP output before performing a callout, unless control =
2585 no_callout_flush is set.
2587 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2588 was true (the default) a successful delivery failed to delete the retry
2589 item, thus causing premature timeout of the address. The bug is now
2592 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2594 PH/21 Long custom messages for fakedefer and fakereject are now split up
2595 into multiline responses in the same way that messages for "deny" and
2596 other ACL rejections are.
2598 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2599 with slight modification.
2601 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2602 draft, changing the syntax and factoring some duplicate code".
2604 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2605 for deliveries of the second and subsequent messages over the same SMTP
2608 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2609 ${reduce, with only minor "tidies".
2611 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2613 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2614 expansion side effects.
2616 PH/27 When a message times out after an over-quota error from an Exim-imposed
2617 quota, the bounce message says "mailbox is full". This message was not
2618 being given when it was a system quota that was exceeded. It now should
2621 MH/03 Made $recipients available in local_scan(). local_scan() already has
2622 better access to the recipient list through recipients_list[], but
2623 $recipients can be useful in postmaster-provided expansion strings.
2625 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2626 in the case of a MAIL command with additional options following the
2627 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2628 were accidentally chopped off.
2630 PH/29 SMTP synchronization checks are implemented when a command is read -
2631 there is a check that no more input is waiting when there shouldn't be
2632 any. However, for some commands, a delay in an ACL can mean that it is
2633 some time before the response is written. In this time, more input might
2634 arrive, invalidly. So now there are extra checks after an ACL has run for
2635 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2636 pipelining has not been advertised.
2638 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2640 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2641 returned in the SMTP response when smtp_return_error_details was false.
2642 This has been fixed.
2644 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2645 instead of the C I/O that was originally supplied, because problems were
2646 reported on Solaris.
2648 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2649 Exim which did not show up earlier: it was assuming that a call to
2650 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2651 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2652 was a macro that became an assignment, so it seemed to work. This has
2653 changed to a proper function call with a void return, hence the compile
2654 error. Exim's code has been fixed.
2656 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2659 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2661 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2663 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2664 because it turns out to be too big or there is a timeout) there is no
2665 "Completed" line in the log. When some messages of this type were
2666 selected by exigrep, they were listed as "not completed". Others were
2667 picked up by some special patterns. I have improved the selection
2668 criteria to be more general.
2670 PH/38 The host_find_failed option in the manualroute router can now be set
2671 to "ignore", to completely ignore a host whose IP address cannot be
2672 found. If all hosts are ignored, the behaviour is controlled by the new
2673 host_all_ignored option.
2675 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2676 homing or because of multiple MX records with /mx) generated more than
2677 one IP address, and the following item turned out to be the local host,
2678 all the secondary addresses of the first item were incorrectly removed
2679 from the list, along with the local host and any following hosts (which
2680 is what is supposed to happen).
2682 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2683 whoever called Exim into the -H file. In the case of the daemon it was
2684 behaving confusingly. When first started, it used values for whoever
2685 started the daemon, but after a SIGHUP it used the Exim user (because it
2686 calls itself on a restart). I have changed the code so that it now always
2689 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2690 message are rejected with the same error (e.g. no authentication or bad
2691 sender address), and a DATA command is nevertheless sent (as can happen
2692 with PIPELINING or a stupid MUA), the error message that was given to the
2693 RCPT commands is included in the rejection of the DATA command. This is
2694 intended to be helpful for MUAs that show only the final error to their
2697 PH/42 Another patch from the Sieve maintainer.
2699 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2700 Eximstats - Fixed some broken HTML links and added missing column headers
2702 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2703 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2705 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2707 PH/43 Yet another patch from the Sieve maintainer.
2709 PH/44 I found a way to check for a TCP/IP connection going away before sending
2710 the response to the final '.' that terminates a message, but only in the
2711 case where the client has not sent further data following the '.'
2712 (unfortunately, this is allowed). However, in many cases there won't be
2713 any further data because there won't be any more messages to send. A call
2714 to select() can be used: if it shows that the input is "ready", there is
2715 either input waiting, or the socket has been closed. An attempt to read
2716 the next input character can distinguish the two cases. Previously, Exim
2717 would have sent an OK response which the client would never have see.
2718 This could lead to message repetition. This fix should cure that, at
2719 least in a lot of common cases.
2721 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2722 advertised in response to EHLO.
2728 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2729 fixed by 4.65/MH/01 (is this a record?) are fixed:
2731 (i) An empty string was always treated as zero by the numeric comparison
2732 operators. This behaviour has been restored.
2734 (ii) It is documented that the numeric comparison operators always treat
2735 their arguments as decimal numbers. This was broken in that numbers
2736 starting with 0 were being interpreted as octal.
2738 While fixing these problems I realized that there was another issue that
2739 hadn't been noticed. Values of message_size_limit (both the global option
2740 and the transport option) were treated as octal if they started with 0.
2741 The documentation was vague. These values are now always treated as
2742 decimal, and I will make that clear in the documentation.
2748 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2749 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2752 MH/01 Don't check that the operands of numeric comparison operators are
2753 integers when their expansion is in "skipping" mode (fixes bug
2754 introduced by 4.64-PH/07).
2756 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2757 child addresses, Exim now panics and dies. Previously, because the count
2758 is held in a short int, deliveries were likely to be lost. As such a
2759 large number of recipients for a single message is ridiculous
2760 (performance will be very, very poor), I have chosen to impose a limit
2761 rather than extend the field.
2767 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2768 leftover -K file (the existence of which was triggered by #402).
2769 While we were at it, introduced process PID as part of the -K
2770 filename. This should rule out race conditions when creating
2773 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2774 processing considerably. Previous code took too long for large mails,
2775 triggering a timeout which in turn triggers #401.
2777 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2778 in the DK code in transports.c. sendfile() is not really portable,
2779 hence the _LINUX specificness.
2781 TF/01 In the add_headers option to the mail command in an Exim filter,
2782 there was a bug that Exim would claim a syntax error in any
2783 header after the first one which had an odd number of characters
2786 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2787 callout verification, Exim cached a "reject" for the entire domain. This
2788 is correct for most verifications, but it is not correct for a recipient
2789 verification with use_sender or use_postmaster set, because in that case
2790 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2791 case of MAIL FROM:<> rejection from other early rejections (e.g.
2792 rejection of HELO). When verifying a recipient using a non-null MAIL
2793 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2794 Whatever the result of the callout, the value of the domain cache is
2795 left unchanged (for any other kind of callout, getting as far as trying
2796 RCPT means that the domain itself is ok).
2798 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2801 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2802 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2803 session, the daemon ignores ECONNECT errors and logs others; it now
2804 ignores EPIPE as well.
2806 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2807 (quoted-printable decoding).
2809 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2810 later the small subsequent patch to fix an introduced bug.
2812 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2814 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2816 PH/08 An error is now given if message_size_limit is specified negative.
2818 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2819 to be given (somewhat) arbitrary names.
2821 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2824 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2825 miscellaneous code fixes
2827 PH/10 Added the log_reject_target ACL modifier to specify where to log
2830 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2831 hostname. This is wrong, because it relates to the incoming message (and
2832 probably the interface on which it is arriving) and not to the outgoing
2833 callout (which could be using a different interface). This has been
2834 changed to use the value of the helo_data option from the smtp transport
2835 instead - this is what is used when a message is actually being sent. If
2836 there is no remote transport (possible with a router that sets up host
2837 addresses), $smtp_active_hostname is used.
2839 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2840 tweaks were necessary in order to get it to work (see also 21 below):
2841 (a) The code assumed that strncpy() returns a negative number on buffer
2842 overflow, which isn't the case. Replaced with Exim's string_format()
2844 (b) There were several signed/unsigned issues. I just did the minimum
2845 hacking in of casts. There is scope for a larger refactoring.
2846 (c) The code used strcasecmp() which is not a standard C function.
2847 Replaced with Exim's strcmpic() function.
2848 (d) The code set only $1; it now sets $auth1 as well.
2849 (e) A simple test gave the error "authentication client didn't specify
2850 service in request". It would seem that Dovecot has changed its
2851 interface. Fortunately there's a specification; I followed it and
2852 changed what the client sends and it appears to be working now.
2854 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2857 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2858 address A is aliased to B and C, where B exists and C does not. Without
2859 -v the output is "A verified" because verification stops after a
2860 successful redirection if more than one address is generated. However,
2861 with -v the child addresses are also verified. Exim was outputting "A
2862 failed to verify" and then showing the successful verification for C,
2863 with its parentage. It now outputs "B failed to verify", showing B's
2864 parentage before showing the successful verification of C.
2866 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2867 look up a TXT record in a specific list after matching in a combined
2870 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2871 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2872 they consult the DNS. I had assumed they would set it the way they
2873 wanted; and indeed my experiments on Linux seem to show that in some
2874 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2875 To be on the safe side, however, I have now made the interface to
2876 host_find_byname() similar to host_find_bydns(), with an argument
2877 containing the DNS resolver options. The host_find_byname() function now
2878 sets these options at its start, just as host_find_bydns() does. The smtp
2879 transport options dns_qualify_single and dns_search_parents are passed to
2880 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2881 of host_find_byname() use the default settings of RES_DEFNAMES
2882 (qualify_single) but not RES_DNSRCH (search_parents).
2884 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2885 spool_read_header() do less string testing, by means of a preliminary
2886 switch on the second character of optional "-foo" lines. (This is
2887 overdue, caused by the large number of possibilities that now exist.
2888 Originally there were few.) While I was there, I also converted the
2889 str(n)cmp tests so they don't re-test the leading "-" and the first
2890 character, in the hope this might squeeze out yet more improvement.
2892 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2893 flag allowing group syntax was set by the header_syntax check but not
2894 turned off, possible causing trouble later; (2) The flag was not being
2895 set at all for the header_verify test, causing "group"-style headers to
2896 be rejected. I have now set it in this case, and also caused header_
2897 verify to ignore an empty address taken from a group. While doing this, I
2898 came across some other cases where the code for allowing group syntax
2899 while scanning a header line wasn't quite right (mostly, not resetting
2900 the flag correctly in the right place). These bugs could have caused
2901 trouble for malformed header lines. I hope it is now all correct.
2903 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2904 with the "reply" argument non-NULL. The code, however (which originally
2905 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2906 but it didn't always do it. This confused somebody who was copying the
2907 code for some other use. I have removed all the tests.
2909 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2910 feature that was used to support insecure browsers during the U.S. crypto
2911 embargo. It requires special client support, and Exim is probably the
2912 only MTA that supported it -- and would never use it because real RSA is
2913 always available. This code has been removed, because it had the bad
2914 effect of slowing Exim down by computing (never used) parameters for the
2915 RSA_EXPORT functionality.
2917 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2918 authenticator to fail if there's a tab character in the incoming data
2919 (there should never be unless someone is messing about, as it's supposed
2920 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2921 if the connection is using TLS or if the remote IP is the same as the
2922 local IP, and the "valid-client-cert option" if a client certificate has
2925 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2926 authenticators. This can be used for authorization after authentication
2927 succeeds. (In the case of plaintext, it servers for both authentication
2930 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2931 if any retry times were supplied.
2933 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2934 connection, where there is no client IP address to check. In this
2935 situation, the verify now always succeeds.
2937 PH/25 Applied John Jetmore's -Mset patch.
2939 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2941 PH/27 In a string expansion for a processed (not raw) header when multiple
2942 headers of the same name were present, leading whitespace was being
2943 removed from all of them, but trailing whitespace was being removed only
2944 from the last one. Now trailing whitespace is removed from each header
2945 before concatenation. Completely empty headers in a concatenation (as
2946 before) are ignored.
2948 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2949 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2951 PH/29 [Removed. This was a change that I later backed out, and forgot to
2952 correct the ChangeLog entry (that I had efficiently created) before
2953 committing the later change.]
2955 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2956 address errors (4xx response to RCPT) over the same connection as other
2957 messages routed to the same hosts. Such deliveries are always "forced",
2958 so retry times are not inspected. This resulted in far too many retries
2959 for the affected addresses. The effect occurred only when there were more
2960 hosts than the hosts_max_try setting in the smtp transport when it had
2961 the 4xx errors. Those hosts that it had tried were not added to the list
2962 of hosts for which the message was waiting, so if all were tried, there
2963 was no problem. Two fixes have been applied:
2965 (i) If there are any address or message errors in an SMTP delivery, none
2966 of the hosts (tried or untried) are now added to the list of hosts
2967 for which the message is waiting, so the message should not be a
2968 candidate for sending over the same connection that was used for a
2969 successful delivery of some other message. This seems entirely
2970 reasonable: after all the message is NOT "waiting for some host".
2971 This is so "obvious" that I'm not sure why it wasn't done
2972 previously. Hope I haven't missed anything, but it can't do any
2973 harm, as the worst effect is to miss an optimization.
2975 (ii) If, despite (i), such a delivery is accidentally attempted, the
2976 routing retry time is respected, so at least it doesn't keep
2977 hammering the server.
2979 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2980 in ${readsocket because some servers need this prod.
2982 PH/32 Added some extra debug output when updating a wait-xxx database.
2984 PH/33 The hint "could be header name not terminated by colon", which has been
2985 given for certain expansion errors for a long time, was not being given
2986 for the ${if def:h_colon_omitted{... case.
2988 PH/34 The spec says: "With one important exception, whenever a domain list is
2989 being scanned, $domain contains the subject domain." There was at least
2990 one case where this was not true.
2992 PH/35 The error "getsockname() failed: connection reset by peer" was being
2993 written to the panic log as well as the main log, but it isn't really
2994 panic-worthy as it just means the connection died rather early on. I have
2995 removed the panic log writing for the ECONNRESET error when getsockname()
2998 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2999 runs only) independently of the message's sender address. This meant
3000 that, if the 4xx error was in fact related to the sender, a different
3001 message to the same recipient with a different sender could confuse
3002 things. In particular, this can happen when sending to a greylisting
3003 server, but other circumstances could also provoke similar problems.
3004 I have changed the default so that the retry time for these errors is now
3005 based a combination of the sender and recipient addresses. This change
3006 can be overridden by setting address_retry_include_sender=false in the
3009 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3010 remote server are returned as part of bounce messages. This was not
3011 happening for LMTP over a pipe (the lmtp transport), but now it is the
3012 same for both kinds of LMTP.
3014 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3015 in header lines that were in fact CNAMEs. This is no longer the case.
3017 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3018 and queue runs started by the daemon processed all messages. This has
3019 been fixed so that -R and -S can now usefully be given with -q<time>.
3021 PH/40 Import PCRE release 6.7 (fixes some bugs).
3023 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3025 PH/42 Give an error if -q is specified more than once.
3027 PH/43 Renamed the variables $interface_address and $interface_port as
3028 $received_ip_address and $received_port, to make it clear that these
3029 values apply to message reception, and not to the outgoing interface when
3030 a message is delivered. (The old names remain recognized, of course.)
3032 PH/44 There was no timeout on the connect() call when using a Unix domain
3033 socket in the ${readsocket expansion. There now is.
3035 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3036 be meaningful with "accept".
3038 SC/01 Eximstats V1.43
3039 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3041 SC/02 Eximstats V1.44
3042 Use a glob alias rather than an array ref in the generated
3043 parser. This improves both readability and performance.
3045 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3046 Collect SpamAssassin and rejection statistics.
3047 Don't display local sender or destination tables unless
3048 there is data to show.
3049 Added average volumes into the top table text output.
3051 SC/04 Eximstats V1.46
3052 Collect data on the number of addresses (recipients)
3053 as well as the number of messages.
3055 SC/05 Eximstats V1.47
3056 Added 'Message too big' to the list of mail rejection
3057 reasons (thanks to Marco Gaiarin).
3059 SC/06 Eximstats V1.48
3060 Mainlog lines which have GMT offsets and are too short to
3061 have a flag are now skipped.
3063 SC/07 Eximstats V1.49 (Alain Williams)
3064 Added the -emptyok flag.
3066 SC/08 Eximstats V1.50
3067 Fixes for obtaining the IP address from reject messages.
3069 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3070 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3071 whitespace changes from 4.64-PH/27
3073 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3076 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3077 are found, allow negative numbers in numeric criteria)
3079 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3081 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3082 to match changes made in 4.64-PH/43
3084 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3086 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3087 "domain" is actually an IP address, and force a failure. This locks out
3088 those revolvers/nameservers that support "A-for-A" lookups, in
3089 contravention of the specifications.
3091 PH/48 When a host name was looked up from an IP address, and the subsequent
3092 forward lookup of the name timed out, the host name was left in
3093 $sender_host_name, contrary to the specification.
3095 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3096 restricted to single-key lookups, Exim was not diagnosing an error if
3097 * or *@ was used with a query-style lookup.
3099 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3101 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3102 long ago, but noone interested enough thought of it. Let's just say that
3103 the "1.1" means that there are some new functions that weren't there at
3104 some point in the past.
3106 PH/51 Error processing for expansion failure of helo_data from an smtp
3107 transport during callout processing was broken.
3109 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3110 tested/used via the -bh/-bhc/-bs options.
3112 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3113 bug, fixed in subsequent PCRE releases).
3115 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3116 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3118 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3124 SC/01 Use a glob alias rather than an array ref in eximstats generated
3125 parser. This improves both readability and performance.
3127 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3128 Don't display local sender or destination tables in eximstats unless
3129 there is data to show.
3130 Added average volumes into the eximstats top table text output.
3132 SC/03 Collect data on the number of addresses (recipients) as well
3133 as the number of messages in eximstats.
3135 TF/01 Correct an error in the documentation for the redirect router. Exim
3136 does (usually) call initgroups() when daemonizing.
3138 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3139 with consistent privilege compared to when running as a daemon.
3141 TF/03 Note in the spec that $authenticated_id is not set for local
3142 submissions from trusted users.
3144 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3145 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3147 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3148 by adding some example configuration directives to the default
3149 configuration file. A little bit of work is required to uncomment the
3150 directives and define how usernames and passwords are checked, but
3151 there is now a framework to start from.
3153 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3154 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3155 without this. I don't know how relevant this is to other LDAP libraries.
3157 PH/02 Add the verb name to the "unknown ACL verb" error.
3159 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3161 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3163 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3164 directory not expanded when it should be if an expanded home directory
3165 was set for the address (which is overridden by the transport).
3167 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3170 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3171 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3172 because it is too late at that time, and has no effect.
3174 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3175 security issue with \' (bugzilla #107). I could not use the
3176 PQescapeStringConn() function, because it needs a PGconn value as one of
3179 PH/08 When testing addresses using -bt, indicate those final addresses that
3180 are duplicates that would not cause an additional delivery. At least one
3181 person was confused, thinking that -bt output corresponded to deliveries.
3182 (Suppressing duplicates isn't a good idea as you lose the information
3183 about possibly different redirections that led to the duplicates.)
3185 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3186 systems where poll() doesn't work, in particular OS X.
3188 PH/10 Added more information to debugging output for retry time not reached.
3190 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3191 operations in malware.c.
3193 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3196 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3197 syslog_duplication set false, log lines that would normally be written
3198 both the the main log and to the reject log were not written to syslog at
3201 PH/14 In the default configuration, change the use of "message" in ACL warn
3202 statements to "add_header".
3204 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3205 not followed by a command (e.g. "seen endif").
3207 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3208 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3211 PH/17 Added extra conditions to the default value of delay_warning_condition
3215 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3216 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3217 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3220 The Auto-Submitted: and various List- headers are standardised, whereas I
3221 don't think Precedence: ever was.
3223 PH/18 Refactored debugging code in route_finduser() to show more information,
3224 in particular, the error code if getpwnam() issues one.
3226 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3227 This is apparently needed in addition to the PH/07 change above to avoid
3228 any possible encoding problems.
3230 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3231 but not after initializing Perl.
3233 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3234 output them only if debugging. By default they are written stderr,
3235 apparently, which is not desirable.
3237 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3240 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3243 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3245 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3246 authenticated or an ident call has been made. Suppress the default
3247 values for $authenticated_id and $authenticated_sender (but permit -oMai
3248 and -oMas) when testing with -bh.
3250 PH/24 Re-jigged the order of the tests in the default configuration so that the
3251 tests for valid domains and recipients precede the DNS black list and CSA
3252 tests, on the grounds that those ones are more expensive.
3254 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3255 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3256 This bug exists in every version of Exim that I still have, right back to
3259 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3260 However, an attempt to turn on case-sensitivity in a regex key by
3261 including (?-i) didn't work because the subject string was already
3262 lowercased, and the effects were non-intuitive. It turns out that a
3263 one-line patch can be used to allow (?-i) to work as expected.
3269 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3270 other effects) broke the use of negated acl sub-conditions.
3272 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3275 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3276 "Deny" causes Exim to reject the incoming connection with a 554 error.
3277 Unfortunately, if there is a major crisis, such as a disk failure,
3278 tcp-wrappers gives "deny", whereas what one would like would be some
3279 kind of temporary error. A kludge has been added to help with this.
3280 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3281 554 error is used if errno is still zero or contains ENOENT (which occurs
3282 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3285 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3287 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3288 errors. Otherwise a message that provokes a temporary error (when other
3289 messages do not) can cause a whole host to time out.
3291 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3292 addresses were routed directly to files or pipes from a redirect router.
3293 File deliveries just didn't batch; pipe deliveries might have suffered
3296 PH/06 A failure to get a lock for a hints database would erroneously always say
3297 "Failed to get write lock", even when it was really a read lock.
3299 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3300 of 0600. This has been changed to use the value of the lockfile_mode
3301 option (which defaults to 0600).
3303 PH/08 Applied small patch from the Sieve maintainer.
3305 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3306 folder from quota calculations, a direct delivery into this folder messed
3307 up the contents of the maildirsize file. This was because the regex was
3308 used only to exclude .Trash (or whatever) when the size of the mailbox
3309 was calculated. There was no check that a delivery was happening into an
3310 excluded directory. This bug has been fixed by ignoring all quota
3311 processing for deliveries into excluded directories.
3313 PH/10 Added the maildirfolder_create_regex option to appendfile.
3319 PH/01 The code for finding all the local interface addresses on a FreeBSD
3320 system running IPv6 was broken. This may well have applied to all BSD
3321 systems, as well as to others that have similar system calls. The broken
3322 code found IPv4 interfaces correctly, but gave incorrect values for the
3323 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3324 that it would not match correctly against @[] and not recognize the IPv6
3327 PH/02 The ipliteral router was not recognizing addresses of the form user@
3328 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3330 PH/03 Added disable_ipv6.
3332 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3333 decoded form, because it is most often used to construct To: headers
3334 lines in autoreplies, and the decoded form may well be syntactically
3335 invalid. However, $reply_address has leading white space removed, and all
3336 newlines turned into spaces so that the autoreply transport does not
3339 PH/05 If group was specified without a user on a router, and no group or user
3340 was specified on a transport, the group from the router was ignored.
3342 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3343 for visible compile-time settings that can be used to change these
3344 numbers, for those that want even more. Backwards compatibility with old
3345 spool files has been maintained. However, going back to a previous Exim
3346 release will lost any variables that are in spool files.
3348 PH/07 Two small changes when running in the test harness: increase delay when
3349 passing a TCP/IP connection to a new process, in case the original
3350 process has to generate a bounce, and remove special handling of
3351 127.0.0.2 (sic), which is no longer necessary.
3353 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3354 be the same on different OS.
3356 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3359 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3360 whether --show-vars was specified or not
3362 JJ/02 exipick: Added support for new ACL variable spool format introduced
3365 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3366 syntactically invalid From: or Reply-to: line, and a filter used this to
3367 generate an autoreply, and therefore failed to obtain an address for the
3368 autoreply, Exim could try to deliver to a non-existent relative file
3369 name, causing unrelated and misleading errors. What now happens is that
3370 it logs this as a hard delivery error, but does not attempt to create a
3373 PH/11 The exinext utility has a -C option for testing purposes, but although
3374 the given file was scanned by exinext itself; it wasn't being passed on
3375 when Exim was called.
3377 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3378 an end-of-file indication when reading a command response.
3380 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3381 compiled. In many other places in Exim, IPv6 addresses are always
3382 recognized, so I have changed this. It also means that IPv4 domain
3383 literals of the form [IPV4:n.n.n.n] are now always recognized.
3385 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3386 used if the router is not running as root, for example, when verifying at
3387 ACL time, or when using -bh. The debugging output from this situation was
3388 non-existent - all you got was a failure to exec. I have made two
3391 (a) Failures to set uid/gid, the current directory, or a process leader
3392 in a subprocess such as that created by queryprogram now generate
3393 suitable debugging output when -d is set.
3395 (b) The queryprogram router detects when it is not running as root,
3396 outputs suitable debugging information if -d is set, and then runs
3397 the subprocess without attempting to change uid/gid.
3399 PH/15 Minor change to Makefile for building test_host (undocumented testing
3402 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3403 additional section of a DNS packet that returns MX or SRV records.
3404 Instead, it always explicitly searches for A/AAAA records. This avoids
3405 major problems that occur when a DNS server includes only records of one
3406 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3407 fixed another bug: if SRV records were looked up and the corresponding
3408 address records were *not* found in the additional section, the port
3409 values from the SRV records were lost.
3411 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3412 using the correct key (the original address) when searching the retry
3413 rules in order to find which one to use for generating the retry hint.
3415 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3416 adding the default one. Similarly, if it contains a Reply-To: header, the
3417 errors_reply_to option, if set, is not used.
3419 PH/19 When calculating a retry time, Exim used to measure the "time since
3420 failure" by looking at the "first failed" field in the retry record. Now
3421 it does not use this if it is later than than the arrival time of the
3422 message. Instead it uses the arrival time. This makes for better
3423 behaviour in cases where some deliveries succeed, thus re-setting the
3424 "first failed" field. An example is a quota failure for a huge message
3425 when small messages continue to be delivered. Without this change, the
3426 "time since failure" will always be short, possible causing more frequent
3427 delivery attempts for the huge message than are intended.
3428 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3430 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3431 $1, $2, $3) because the numerical variables can be reset during some
3432 expansion items (e.g. "match"), thereby losing the authentication data.
3434 PH/21 Make -bV show the size of off_t variables so that the test suite can
3435 decide whether to run tests for quotas > 2G.
3437 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3438 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3439 filecount value is greater than 2G or if a quota value is greater than 2G
3440 on a system where the size of off_t is not greater than 4, a panic error
3443 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3444 never match. The debug and -bh output now contains an explicit error
3445 message indicating a malformed IPv4 address or mask.
3447 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3448 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3449 PH/23 above applies.
3451 PH/25 Do not write to syslog when running in the test harness. The only
3452 occasion when this arises is a failure to open the main or panic logs
3453 (for which there is an explicit test).
3455 PH/26 Added the /no_tell option to "control=freeze".
3457 PH/27 If a host name lookup failed very early in a connection, for example, if
3458 the IP address matched host_lookup and the reverse lookup yielded a name
3459 that did not have a forward lookup, an error message of the form "no IP
3460 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3461 could be logged. Now it outputs the IP address instead of "NULL".
3463 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3464 allows the sender and the authenticated sender to be set when
3465 submitting a message from within Exim. Since child_open_exim() is
3466 documented for local_scan(), the new function should be too.
3468 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3469 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3470 results in an empty string is now treated as unset.
3472 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3474 PH/31 Added sender_verify_fail logging option.
3476 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3477 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3478 tidied the source and removed it altogether.
3480 PH/33 When a queue run was abandoned because the load average was too high, a
3481 log line was always written; now it is written only if the queue_run log
3482 selector is set. In addition, the log line for abandonment now contains
3483 information about the queue run such as the pid. This is always present
3484 in "start" and "stop" lines but was omitted from the "abandon" line.
3486 PH/34 Omit spaces between a header name and the colon in the error message that
3487 is given when verify = headers_syntax fails (if there are lots of them,
3488 the message gets confusing).
3490 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3491 names, as there are now some PTR records that contain slashes. This check
3492 is only to protect against broken name servers that fall over on strange
3493 characters, so the fact that it applies to all lookups doesn't matter.
3495 PH/36 Now that the new test suite is complete, we can remove some of the
3496 special code in Exim that was needed for the old test suite. For example,
3497 sorting DNS records because real resolvers return them in an arbitrary
3498 order. The new test suite's fake resolver always returns records in the
3501 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3502 bounces) except when queue_only is set, to avoid logging races between
3503 the different processes.
3505 PH/38 Panic-die if .include specifies a non-absolute path.
3507 PH/39 A tweak to the "H" retry rule from its user.
3509 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3510 a label. They prevented compilation on older perls.
3512 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3513 a warning to be raised on newish perls.
3515 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3516 on queue. Changes to match documented behaviour of showing count of
3517 messages matching specified criteria.
3519 PH/40 Changed the default ident timeout from 30s to 5s.
3521 PH/41 Added support for the use of login_cap features, on those BSD systems
3522 that have them, for controlling the resources used by pipe deliveries.
3524 PH/42 The content-scanning code uses fopen() to create files in which to put
3525 message data. Previously it was not paying any attention to the mode of
3526 the files. Exim runs with umask(0) because the rest of the code creates
3527 files with open(), and sets the required mode explicitly. Thus, these
3528 files were ending up world-writeable. This was not a big issue, because,
3529 being within the spool directory, they were not world-accessible. I have
3530 created a function called modefopen, which takes an additional mode
3531 argument. It sets umask(777), creates the file, chmods it to the required
3532 mode, then resets the umask. All the relevant calls to fopen() in the
3533 content scanning code have been changed to use this function.
3535 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3536 to 24 hours. This avoids potential overflow problems when processing G
3537 and H retry rules. I suspect nobody ever tinkers with this value.
3539 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3541 PH/45 When the plaintext authenticator is running as a client, the server's
3542 challenges are checked to ensure they are valid base64 strings. By
3543 default, the authentication attempt is cancelled if an invalid string is
3544 received. Setting client_ignore_invalid_base64 true ignores these errors.
3545 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3546 they are received. Thus, the responses can be made to depend on the
3547 challenges. If an invalid string is ignored, an empty string is placed in
3550 PH/46 Messages that are created by the autoreply transport now contains a
3551 References: header, in accordance with RFCs 2822 and 3834.
3553 PH/47 Added authenticated_sender_force to the smtp transport.
3555 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3557 PH/49 Installed latest patch from the Sieve maintainer.
3559 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3560 was also set, the appendfile transport was unnecessarily scanning a
3561 directory of message files (e.g. for maildir delivery) to find the count
3562 of files (along with the size), even though it did not need this
3563 information. It now does the scan only if it needs to find either the
3564 size of the count of files.
3566 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3568 PH/52 Two bugs concerned with error handling when the smtp transport is
3571 (i) Exim was not creating retry information for temporary errors given
3572 for individual recipients after the DATA command when the smtp transport
3573 was used in LMTP mode. This meant that they could be retried too
3574 frequently, and not timed out correctly.
3576 (ii) Exim was setting the flag that allows error details to be returned
3577 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3578 recipients that were returned after the DATA command.
3580 PH/53 This is related to PH/52, but is more general: for any failing address,
3581 when detailed error information was permitted to be returned to the
3582 sender, but the error was temporary, then after the final timeout, only
3583 "retry timeout exceeded" was returned. Now it returns the full error as
3584 well as "retry timeout exceeded".
3586 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3587 do this, and (what is worse) MTAs that accept it.
3589 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3590 will now be deprecated.
3592 PH/56 New os.c-cygwin from the Cygwin maintainer.
3594 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3595 formats (previously only available in exim formats via -bpr, -bpru,
3596 and -bpra. Now also available in native and exiqgrep formats)
3598 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3599 with very large, slow to parse queues
3601 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3603 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3605 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3606 responses to authentication challenges, though it was showing the
3607 challenges; (ii) I've removed the CR characters from the debug output for
3610 PH/58 Allow for the insertion of a newline as well as a space when a string
3611 is turned into more than one encoded-word during RFC 2047 encoding. The
3612 Sieve code now uses this.
3614 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3615 data_4xx, lost_connection, tls_required.
3617 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3618 message was being sent as an SMTP response.
3620 PH/61 Add -l and -k options to exicyclog.
3622 PH/62 When verifying, if an address was redirected to one new address, so that
3623 verification continued, and the new address failed or deferred after
3624 having set something in $address_data, the value of $address_data was not
3625 passed back to the ACL. This was different to the case when no
3626 redirection occurred. The value is now passed back in both cases.
3628 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3629 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3630 use login_cap.h, so on its own it isn't the distinguishing feature. The
3631 new name refers directly to the setclassresources() function.
3633 PH/65 Added configuration files for NetBSD3.
3635 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3637 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3638 is preferred over IPv4.
3640 PH/68 The bounce_return_message and bounce_return_body options were not being
3641 honoured for bounces generated during the reception of non-SMTP messages.
3642 In particular, this applied to messages rejected by the ACL. This bug has
3643 been fixed. However, if bounce_return_message is true and bounce_return_
3644 body is false, the headers that are returned for a non-SMTP message
3645 include only those that have been read before the error was detected.
3646 (In the case of an ACL rejection, they have all been read.)
3648 PH/69 The HTML version of the specification is now built in a directory called
3649 spec_html instead of spec.html, because the latter looks like a path with
3650 a MIME-type, and this confuses some software.
3652 PH/70 Catch two compiler warnings in sieve.c.
3654 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3655 function verify_get_ident() calls ip_connect() to connect a socket, but
3656 if the "connect()" function timed out, ip_connect() used to close the
3657 socket. However, verify_get_ident() also closes the socket later, and in
3658 between Exim writes to the log, which may get opened at this point. When
3659 the socket was closed in ip_connect(), the log could get the same file
3660 descriptor number as the socket. This naturally causes chaos. The fix is
3661 not to close the socket in ip_connect(); the socket should be closed by
3662 the function that creates it. There was only one place in the code where
3663 this was missing, in the iplookup router, which I don't think anybody now
3664 uses, but I've fixed it anyway.
3666 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3667 well as to direct DNS lookups. Otherwise the handling of names in host
3668 lists is inconsistent and therefore confusing.
3674 PH/01 Two changes to the default runtime configuration:
3676 (1) Move the checks for relay_from_hosts and authenticated clients from
3677 after to before the (commented out) DNS black list checks.
3679 (2) Add control=submission to the relay_from_hosts and authenticated
3680 clients checks, on the grounds that messages accepted by these
3681 statements are most likely to be submissions.
3683 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3685 (1) Generate an error if the third argument for the ${prvs expansion is
3688 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3691 (3) Reset the variables that are obtained from the first argument of
3692 ${prvscheck and used in the second argument before leaving the code,
3693 because their memory is reclaimed, so using them afterwards may do
3696 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3697 one (it's much easier than Tom thought :-).
3699 (5) Because of (4), we can now allow for the use of $prvscheck_result
3700 inside the third argument.
3702 PH/03 For some reason, the default setting of PATH when running a command from
3703 a pipe transport was just "/usr/bin". I have changed it to
3706 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3707 anything to be listed in the output from -bV.
3709 PH/05 When a filter generated an autoreply, the entire To: header line was
3710 quoted in the delivery log line, like this:
3712 => >A.N.Other <ano@some.domain> <original@ddress> ...
3714 This has been changed so that it extracts the operative address. There
3715 may be more than one such address. If so, they are comma-separated, like
3718 => >ano@some.domain,ona@other.domain <original@ddress> ...
3720 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3721 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3722 not being looked up in the rDNS to get a host name, Exim was showing the
3723 IP address twice in Received: lines, even though the IP addresses were
3724 identical. For example:
3726 Received: from [1.2.3.4] (helo=[1.2.3.4])
3728 However, if the real host name was known, it was omitting the HELO data
3729 if it matched the actual IP address. This has been tidied up so that it
3730 doesn't show the same IP address twice.
3732 PH/07 When both +timestamp and +memory debugging was on, the value given by
3733 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3734 called by the debug printing, thereby overwriting the timestamp buffer.
3735 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3737 PH/08 When the original message was included in an autoreply transport, it
3738 always said "this is a copy of the message, including all the headers",
3739 even if body_only or headers_only was set. It now gives an appropriate
3742 PH/09 Applied a patch from the Sieve maintainer which:
3744 o fixes some comments
3745 o adds the (disabled) notify extension core
3746 o adds some debug output for the result of if/elsif tests
3747 o points to the current vacation draft in the documentation
3748 and documents the missing references header update
3752 o fixes a bug in processing the envelope test (when testing
3753 multiple envelope elements, the last element determined the
3756 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3757 Electronic Mail") by including:
3759 Auto-submitted: auto-generated
3761 in the messages that it generates (bounce messages and others, such as
3762 warnings). In the case of bounce messages for non-SMTP messages, there was
3763 also a typo: it was using "Auto_submitted" (underscore instead of
3764 hyphen). Since every message generated by Exim is necessarily in response
3765 to another message, thes have all been changed to:
3767 Auto-Submitted: auto-replied
3769 in accordance with these statements in the RFC:
3771 The auto-replied keyword:
3773 - SHOULD be used on messages sent in direct response to another
3774 message by an automatic process,
3776 - MUST NOT be used on manually-generated messages,
3778 - MAY be used on Delivery Status Notifications (DSNs) and Message
3779 Disposition Notifications (MDNs),
3781 - MUST NOT be used on messages generated by automatic or periodic
3782 processes, except for messages which are automatic responses to
3785 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3786 to the default Received: header definition.
3788 PH/12 Added log selector acl_warn_skipped (default on).
3790 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3791 variables because (a) they are in the wrong storage pool and (b) even if
3792 they were copied, it wouldn't work properly because of the caching.
3794 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3795 checking when decoding. Apparently there are clients that generate
3796 overlong encoded strings. Why am I not surprised?
3798 PH/15 If the first argument of "${if match_address" was not empty, but did not
3799 contain an "@" character, Exim crashed. Now it writes a panic log message
3800 and treats the condition as false.
3802 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3804 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3805 extension "envelope-auth". The code is finished and in agreement with
3806 other implementations, but there is no documentation so far and in fact,
3807 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3808 not changing the active code.
3810 Print executed "if" and "elsif" statements when debugging is used. This
3811 helps a great deal to understand what a filter does.
3813 Document more things not specified clearly in RFC3028. I had all this
3814 sorted out, when out of a sudden new issues came to my mind. Oops."
3816 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3819 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3820 canonical form (as documented). However, after a host name lookup from
3821 the IP address, check_host() was doing a simple string comparison with
3822 addresses acquired from the DNS when checking that the found name did
3823 have the original IP as one of its addresses. Since any found IPv6
3824 addresses are likely to be in abbreviated form, the comparison could
3825 fail. Luckily, there already exists a function for doing the comparison
3826 by converting both addresses to binary, so now that is used instead of
3827 the text comparison.
3829 PH/20 There was another similar case to PH/19, when a complete host name was
3830 given in a host list; looking up its IP address could give an abbreviated
3831 form, whereas the current host's name might or might not be abbreviated.
3832 The same fix has been applied.
3838 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3839 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3842 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3843 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3845 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3847 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3848 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3849 Also, for "Auto-Submitted", treat anything other than "no" as
3850 non-personal, in accordance with RFC 3834. (Previously it treated
3851 anything starting "auto-" as non-personal.)
3853 TF/01 The control=submission/name=... option had a problem with syntax
3854 errors if the name included a slash character. The /name= option
3855 now slurps the rest of the string, so it can include any characters
3856 but it must come last in the list of options (after /sender_retain
3859 PH/05 Some modifications to the interface to the fake nameserver for the new
3867 TK/01 Added the "success_on_redirect" address verification option. See
3868 NewStuff for rationale and an example.
3870 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3872 PH/02 Patch to exigrep to allow it to work on syslog lines.
3874 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3875 fread() to skip over the body file's header line, because in Cygwin the
3876 header line is locked and is inaccessible.
3878 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3879 co-exist for some time) to make it clear that it is the Exim ID that is
3880 referenced, not the Message-ID: header line.
3882 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3883 string_format() function, because snprintf() does not exist on all
3886 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3887 ${stat: expansion item.
3889 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3890 protocol synchronization error", to keep the pedants happy.
3892 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3893 well as for IRIX systems, when gcc is being used. See the host.c source
3896 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3898 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3901 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3902 transport and to the smtp transport in LMTP mode.
3904 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3906 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3907 run a filter in a subprocess. This could lead to confusion in subsequent
3908 lookups in the parent process. There should also be a search_tidyup() at
3909 the end of the subprocess.
3911 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3912 only if the host matched helo_try_verify_hosts, which caused the
3913 verification to occur when the EHLO/HELO command was issued. The ACL just
3914 tested the remembered result. Now, if a previous verification attempt has
3915 not happened, "verify = helo" does it there and then.
3917 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3919 TK/03 Fix log output including CR from clamd.
3921 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3922 exist provoked a memory error which could cause a segfault.
3924 PH/15 Installed PCRE 6.2
3926 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3928 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3929 of the problem. Specifically, suggested +O2 rather than +O1 for the
3932 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3934 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3935 record, the port was not correctly logged when the outgoing_port log
3936 selector was set (it logged the transort's default port).
3938 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3939 fallback_hosts, and "hosts" in the smtp transport.
3941 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3942 host errors such as "Connection refused".
3944 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3945 authentication with radiusclient 0.4.9:
3947 - Error returned from rc_read_config was caught wrongly
3948 - Username/password not passed on to radius server due to wrong length.
3950 The presumption is that some radiusclient API changes for 4.51/PH/17
3951 were not taken care of correctly. The code is still untested by me (my
3952 Linux distribution still has 0.3.2 of radiusclient), but it was
3953 contributed by a Radius user.
3955 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3956 expanding the "port" option of the smtp transport.
3958 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3959 while reading a MIME header. Thanks to Tom Hughes for a patch.
3961 PH/24 Include config.h inside local_scan.h so that configuration settings are
3964 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3965 that in an ACL for RCPT (for example), you can examine exactly what was
3968 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3969 commands, but it was not correctly comparing the address with the actual
3970 client host address. Thus, it would show the EHLO address in Received:
3971 header lines when this was not necessary.
3973 PH/27 Added the % operator to ${eval:}.
3975 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3976 it should be ignoring failures (because with -C, for example, it has lost
3977 privilege). It wasn't ignoring creation failures other than "already
3980 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3983 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3984 because an input line was too long, either on its own, or by virtue of
3985 too many continuations, the temporary file was not being removed, and the
3986 return code was incorrect.
3988 PH/31 Missing "BOOL" in function definition in filtertest.c.
3990 PH/32 Applied Sieve patches from the maintainer.
3992 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3994 PH/33 Added "verify = not_blind".
3996 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3997 Local/Makefile (with some defaults set). These are used in built scripts
3998 such as exicyclog, but they have never been used in the exim_install
3999 script (though there are many overriding facilities there). I have
4000 arranged that the exim_install script now takes note of these two
4003 PH/35 Installed configuration files for Dragonfly.
4005 PH/36 When a locally submitted message by a trusted user did not contain a
4006 From: header, and the sender address was obtained from -f or from an SMTP
4007 MAIL command, and the trusted user did not use -F to supply a sender
4008 name, $originator_name was incorrectly used when constructing a From:
4009 header. Furthermore, $originator_name was used for submission mode
4010 messages from external hosts without From: headers in a similar way,
4011 which is clearly wrong.
4013 PH/37 Added control=suppress_local_fixups.
4015 PH/38 When log_selector = +received_sender was set, and the addition of the
4016 sender made the log line's construction buffer exactly full, or one byte
4017 less than full, an overflow happened when the terminating "\n" was
4020 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4021 when the result of a list match is failure because a DNS lookup failed.
4023 PH/40 RM_COMMAND is now used in the building process.
4025 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4026 the "build-* directories that it finds.
4028 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4029 address in a domain literal was a prefix of an interface address.
4031 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4032 when verifying a sender address, unless rewrite_headers is false.
4034 PH/44 Wrote a long comment about why errors_to addresses are verified as
4035 recipients, not senders.
4037 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4038 the ratelimit ACL was added.
4040 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4042 PH/46 Added extra information about PostgreSQL errors to the error string.
4044 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4045 avoiding the need to install special zones in a real server. This is
4046 backwards compatible; if it can't find the fake resolver, it drops back.
4047 Thus, both old and new test suites can be run.
4049 TF/02 Added util/ratelimit.pl
4051 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4052 clock is set back in time.
4054 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4055 Candler <B.Candler@pobox.com>.
4057 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4058 OK for addresses that are the result of redirections.
4060 PH/48 A number of further additions for the benefit of the new test suite,
4061 including a fake gethostbyname() that interfaces to the fake DNS resolver
4064 TF/06 The fix for widen_domains has also been applied to qualify_single and
4065 search_parents which are the other dnslookup options that can cause
4068 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4071 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4073 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4074 service" if it encountered an MX record with an empty target hostname.
4075 The message is now "an MX or SRV record indicated no SMTP service".
4077 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4078 if the DNS of the sending site is misconfigured. This is quite a
4079 common situation. This change restores the behaviour of treating a
4080 helo verification defer as a failure.
4082 PH/51 If self=fail was set on a router, the bounce message did not include the
4083 actual error message.
4089 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4091 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4092 command itself ended in error, the underlying message about the transport
4093 filter timeout was being overwritten with the pipe command error. Now the
4094 underlying error message should be appended to the second error message.
4096 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4098 PH/02 Reduce the amount of output that "make" produces by default. Full output
4099 can still be requested.
4101 PH/03 The warning log line about a condition test deferring for a "warn" verb
4102 was being output only once per connection, rather than after each
4103 occurrence (because it was using the same function as for successful
4104 "warn" verbs). This seems wrong, so I have changed it.
4106 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4107 it should not have, which might have caused a crash in the right
4108 circumstances, but probably never did.
4110 PH/04 Installed a modified version of Tony Finch's patch to make submission
4111 mode fix the return path as well as the Sender: header line, and to
4112 add a /name= option so that you can make the user's friendly name appear
4115 TF/03 Added the control = fakedefer ACL modifier.
4117 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4118 Mark Lowes for thorough testing.
4120 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4122 TK/03 Merged latest SRS patch from Miles Wilton.
4124 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4125 with the definition in sysexits.h (which is #included earlier).
4126 Fortunately, Exim does not actually use EX_OK. The code used to try to
4127 preserve the sysexits.h value, by assuming that macro definitions were
4128 scanned for macro replacements. I have been disabused of this notion,
4129 so now the code just undefines EX_OK before #including unistd.h.
4131 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4132 in the smtp transport. When a block could not be written in a single
4133 write() function, the timeout was being re-applied to each part-write.
4134 This seems wrong - if the receiver was accepting one byte at a time it
4135 would take for ever. The timeout is now adjusted when this happens. It
4136 doesn't have to be particularly precise.
4138 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4139 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4141 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4142 without a domain if the check to <postmaster@domain> fails.
4144 SC/01 Eximstats: added -xls and the ability to specify output files
4145 (patch written by Frank Heydlauf).
4147 SC/02 Eximstats: use FileHandles for outputting results.
4149 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4151 SC/04 Eximstats: fixed display of large numbers with -nvr option
4153 SC/05 Eximstats: fixed merging of reports with empty tables.
4155 SC/06 Eximstats: added the -include_original_destination flag
4157 SC/07 Eximstats: removed tabs and trailing whitespace.
4159 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4161 TK/06 MBOX spool code: Add real "From " MBOX separator line
4162 so the .eml file is really in mbox format (even though
4163 most programs do not really care). Patch from Alex Miller.
4165 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4166 The latter is generated from $received_to and is only set if the
4167 message has one envelope recipient. SA can use these headers,
4168 obviously out-of-the-box. Patch from Alex Miller.
4170 PH/08 The ${def test on a variable was returning false if the variable's
4171 value was "0", contrary to what the specification has always said!
4172 The result should be true unless the variable is empty.
4174 PH/09 The syntax error of a character other than { following "${if
4175 def:variable_name" (after optional whitespace) was not being diagnosed.
4176 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4177 accidental colon was present, for example, could give incorrect results.
4179 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4180 result is used (not including appendfile, where other changes are about
4183 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4184 This involved changing a lot of size variables from int to off_t. It
4185 should work with maildirs and everything.
4187 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4188 spamd dying while we are connected to it.
4190 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4193 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4194 function for BDB 4.3.
4196 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4198 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4199 the sources for PCRE have been re-arranged and more files are now
4202 PH/15 The code I had for printing potentially long long variables in PH/11
4203 above was not the best (it lost precision). The length of off_t variables
4204 is now inspected at build time, and an appropriate printing format (%ld
4205 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4206 to be "long long int" or "long int". This is needed for the internal
4207 formatting function string_vformat().
4209 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4210 the configuration file to be ":syslog", then the script "guesses" where
4211 the logs files are, rather than using the compiled in default. In our
4212 case the guess is not the same as the compiled default, so the script
4213 suddenly stopped working when I started to use syslog. The patch checks
4214 to see if log_file_path is "". If so, it attempts to read it from exim
4215 with no configuration file to get the compiled in version, before it
4216 falls back to the previous guessing code."
4218 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4219 implementing BATV in an Exim configuration. See NewStuff for the gory
4222 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4223 Makefile that are specific to HP-UX.
4225 PH/18 If the "use_postmaster" option was set for a recipient callout together
4226 with the "random" option, the postmaster address was used as the MAIL
4227 FROM address for the random test, but not for the subsequent recipient
4228 test. It is now used for both.
4230 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4231 patch removes a few documentation additions to RFC 3028, because the
4232 latest draft now contains them. It adds the new en;ascii-case comparator
4233 and a new error check for 8bit text in MIME parts. Comparator and
4234 require names are now matched exactly. I enabled the subaddress
4235 extension, but it is not well tested yet (read: it works for me)."
4237 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4238 rework some of the code of TK/09 above to avoid the hardwired use of
4239 "%lld" and "long long". Replaced the call to snprintf() with a call to
4242 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4243 records point to non-existent hosts", "retry timeout exceeded", and
4244 "retry time not reached for any host after a long failure period".
4246 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4247 experimental DomainKeys support:
4249 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4250 (2) On an error such as an illegally used "control", the wrong name for
4251 the control was given.
4253 These problems did NOT occur unless DomainKeys support was compiled.
4255 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4257 PH/24 Added ${if match_ip condition.
4259 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4260 databases so that it will be absolutely obvious if a crash occurs in the
4261 DB library. This is a regular occurrence (often caused by mis-matched
4264 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4265 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4266 picked up on a user's system that detects such things. There doesn't seem
4267 to be a gcc warning option for this - only an attribute that has to be
4268 put on the function's prototype. It seems that in Fedora Core 4 they have
4269 set this on a number of new functions. No doubt there will be more in due
4272 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4273 specify a transport. However, if an address that was verified by such a
4274 router was the subject of a callout, Exim crashed because it tried to
4275 read the rcpt_include_affixes from the non-existent transport. Now it
4276 just assumes that the setting of that option is false. This bug was
4277 introduced by 4.51/PH/31.
4279 PH/28 Changed -d+all to exclude +memory, because that information is very
4280 rarely of interest, but it makes the output a lot bigger. People tend to
4281 do -d+all out of habit.
4283 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4284 code in os-type was giving problems when libc.so lives in lib64, like on
4287 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4288 aren't the modern standard, and it seems that some systems' include files
4289 don't always have them. Exim was already checking for some of the newer
4290 ones like T_AAAA, and defining it itself. I've added checks for all the
4291 record types that Exim uses.
4293 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4294 not automatically generating a new one, as it is supposed to. This
4295 prevented TLS from working. If the file did exist, but contained invalid
4296 data, a new version was generated, as expected. It was only the case of a
4297 non-existent file that was broken.
4299 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4300 with a change in libdomainkeys > 0.64.
4302 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4303 from DNS. If the selector record carries the flag, it now has
4304 precedence over the domain-wide flag.
4306 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4308 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4309 the use of a port name that isn't defined in /etc/services) occurred, the
4310 message was deferred as in a normal delivery, and thus remained on the
4311 spool, instead of being failed because of the mua_wrapper setting. This
4312 is now fixed, and I tidied up some of the mua_wrapper messages at the
4315 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4316 the messages in a hash of arrays rather than using individual hashes.
4317 This is a bit cleaner and results in dramatic memory savings, albeit
4318 at a slight CPU cost.
4320 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4321 as requested by Marc Sherman.
4323 SC/10 Eximstats: added histograms for user specified patterns as requested
4326 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4328 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4329 fopen() in the content-scanning modules that did not already have it.
4335 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4336 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4338 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4340 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4342 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4343 to test Sieve filters that use "vacation".
4345 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4346 that changes the way the GnuTLS parameters are stored in the cache file.
4347 The new format can be generated externally. For backward compatibility,
4348 if the data in the cache doesn't make sense, Exim assumes it has read an
4349 old-format file, and it generates new data and writes a new file. This
4350 means that you can't go back to an older release without removing the
4353 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4354 work if there are any delivery delays because "one_time" forces the
4355 parent to be marked "delivered", so its unseen clone is never tried
4356 again. For this reason, Exim now forbids the simultaneous setting of
4359 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4360 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4361 entry if you want to know the details. The fix, however, neglected to
4362 consider the case when local delivery batching is involved. The test for
4363 "previously delivered" was not happening when checking to see if an
4364 address could be batched with a previous (undelivered) one; under
4365 certain circumstances this could lead to multiple deliveries to the same
4368 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4369 in its include files, and this causes problems building Exim.
4371 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4372 header_syntax) but Exim was just ignoring anything given after a slash.
4373 In particular, this caused confusion with an attempt to use "verify =
4374 reverse_host_lookup/defer_ok". An error is now given when options are
4375 supplied for verify items that do not have them. (Maybe reverse_host_
4376 lookup should have a defer_ok option, but that's a different point.)
4378 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4379 defined by RFC 821) to 2048, because there were problems with some AUTH
4380 commands, and RFC 1869 says the size should be increased for extended
4381 SMTP commands that take arguments.
4383 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4386 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4387 "unknown" error; now it says that the functionality isn't in the binary.
4389 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4390 an address' error message when a string expansion fails (syntax or
4391 whatever). Otherwise the password may appear in the log. Following change
4392 PH/42 below, there is no longer a chance of it appearing in a bounce
4395 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4397 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4398 of its addresses was ever tried. (Bugzilla bug #2.)
4400 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4401 the result incorrectly in the debug output. (It correctly added a newline
4402 to what was transported.)
4404 TF/01 Added $received_time.
4406 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4407 commented out examples of how to interface to a virus scanner and to
4408 SpamAssassin. Also added commented examples of av_scanner and
4409 spamd_address settings.
4411 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4412 and controls are allowed in which ACLs. There were a couple of minor
4413 errors. Some of the entries in the conditions table (which is a table of
4414 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4415 negation of where the condition IS allowed.
4417 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4419 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4420 header file does not have a version number, so I've had to invent a new
4421 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4422 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4423 radiusclient), but it was contributed by a Radius user.
4425 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4426 files or directories, for OpenSSL.
4428 PH/19 When an Exim process that is running as root has to create an Exim log
4429 file, it does so in a subprocess that runs as exim:exim so as to get the
4430 ownership right at creation (otherwise, other Exim processes might see
4431 the file with the wrong ownership). There was no test for failure of this
4432 fork() call, which would lead to the process getting stuck as it waited
4433 for a non-existent subprocess. Forks do occasionally fail when resources
4434 run out. I reviewed all the other calls to fork(); they all seem to check
4437 PH/20 When checking for unexpected SMTP input at connect time (before writing
4438 the banner), Exim was not dealing correctly with a non-positive return
4439 from the read() function. If the client had disconnected by this time,
4440 the result was a log entry for a synchronization error with an empty
4441 string after "input=" when read() returned zero. If read() returned -1
4442 (an event I could not check), uninitialized data bytes were printed.
4443 There were reports of junk text (parts of files, etc) appearing after
4446 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4448 PH/22 Added support for macro redefinition, and (re)definition in between
4449 driver and ACL definitions.
4451 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4452 forgetting to use the resulting value; it was using the unexpanded value.
4454 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4455 hadn't been configured. The fix is from Juergen Kreileder, who
4456 understands it better than I do:
4458 "Here's what I see happening with three configured cyrus_sasl
4459 authenticators configured (plain, login, cram-md5):
4461 On startup auth_cyrus_sasl_init() gets called for each of these.
4462 This means three calls to sasl_listmech() without a specified mech_list.
4463 => SASL tests which mechs of all available mechs actually work
4464 => three warnings about OTP not working
4465 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4467 With the patch, sasl_listmech() also gets called three times. But now
4468 SASL's mech_list option is set to the server_mech specified in the the
4469 authenticator. Or in other words, the answer from sasl_listmech()
4470 gets limited to just the mech you're testing for (which is different
4472 => the return list contains just 'plain' or 'login', 'cram-md5' or
4473 nothing depending on the value of ob->server_mech.
4475 I've just tested the patch: Authentication still works fine,
4476 unavailable mechs specified in the exim configuration are still
4477 caught, and the auth.log warnings about OTP are gone."
4479 PH/25 When debugging is enabled, the contents of the command line are added
4480 to the debugging output, even when log_selector=+arguments is not
4483 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4484 answer is "GNU", and only if the return is "GNU/something" is the answer
4487 PH/27 $acl_verify_message is now set immediately after the failure of a
4488 verification in an ACL, and so is available in subsequent modifiers. In
4489 particular, the message can be preserved by coding like this:
4491 warn !verify = sender
4492 set acl_m0 = $acl_verify_message
4494 Previously, $acl_verify_message was set only while expanding "message"
4495 and "log_message" when a very denied access.
4497 PH/28 Modified OS/os.c-Linux with
4499 -#ifndef OS_LOAD_AVERAGE
4500 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4502 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4503 nomenclature these days.)
4505 PH/29 Installed patch from the Sieve maintainer that adds the options
4506 sieve_useraddress and sieve_subaddress to the redirect router.
4508 PH/30 In these circumstances:
4509 . Two addresses routed to the same list of hosts;
4510 . First host does not offer TLS;
4511 . First host accepts first address;
4512 . First host gives temporary error to second address;
4513 . Second host offers TLS and a TLS session is established;
4514 . Second host accepts second address.
4515 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4516 and peerdn, if requested) that were in fact used only for the second
4519 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4520 attention to any local part prefix or suffix that was matched by the
4521 router that accepted the address. It now behaves in the same way as it
4522 does for delivery: the affixes are removed from the local part unless
4523 rcpt_include_affixes is set on the transport.
4525 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4526 timeout during the DATA phase of an incoming message.
4528 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4529 applied a patch sanctioned by the Sieve maintainer.
4531 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4532 the uid or gid is negative. A case of a negative gid caused this to be
4533 noticed. The fix allows for either to be negative.
4535 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4536 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4539 PH/36 The change PH/12 above was broken. Fixed it.
4541 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4542 the grounds that routing the same address twice would always produce the
4543 same answer. This might have been true once, but it is certainly no
4544 longer true now. Routing a child address may depend on the previous
4545 routing that produced that child. Some complicated redirection strategies
4546 went wrong when messages had multiple recipients, and made Exim's
4547 behaviour dependent on the order in which the addresses were given.
4549 I have moved the duplicate checking until after the routing is complete.
4550 Exim scans the addresses that are assigned to local and remote
4551 transports, and removes any duplicates. This means that more work will be
4552 done, as duplicates will always all be routed, but duplicates are
4553 presumably rare, so I don't expect this is of any significance.
4555 For deliveries to pipes, files, and autoreplies, the duplicate checking
4556 still happens during the routing process, since they are not going to be
4559 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4560 It corrects a timeout issue with spamd. This is Ian's comment: "The
4561 background is that sometimes spamd either never reads data from a
4562 connection it has accepted, or it never writes response data. The exiscan
4563 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4564 blindly assumes that writes won't block so it may never time out."
4566 PH/39 Allow G after quota size as well as K and M.
4568 PH/40 The value set for $authenticated_id in an authenticator may not contain
4569 binary zeroes or newlines because the value is written to log lines and
4570 to spool files. There was no check on this. Now the value is run through
4571 the string_printing() function so that such characters are converted to
4572 printable escape sequences.
4574 PH/41 $message_linecount is a new variable that contains the total number of
4575 lines in the message. Compare $body_linecount, which is the count for the
4578 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4579 bounce and delay warning messages, except in certain special cases, which
4582 (a) An SMTP error message from a remote host;
4583 (b) A message specified in a :fail: redirection;
4584 (c) A message specified in a "fail" command in a system filter;
4585 (d) A message specified in a FAIL return from the queryprogram router;
4586 (e) A message specified by the cannot_route_message router option.
4588 In these cases only, Exim does include the error details in bounce and
4589 warning messages. There are also a few cases where bland messages such
4590 as "unrouteable address" or "local delivery error" are given.
4592 PH/43 $value is now also set for the "else" part of a ${run expansion.
4594 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4595 being worked on, but at least Exim now implements the latest version to
4598 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4599 process to complete was treated as a delivery failure, a timeout while
4600 writing the message to the pipe was logged, but erroneously treated as a
4601 successful delivery. Such timeouts include transport filter timeouts. For
4602 consistency with the overall process timeout, these timeouts are now
4603 treated as errors, giving rise to delivery failures by default. However,
4604 there is now a new Boolean option for the pipe transport called
4605 timeout_defer, which, if set TRUE, converts the failures into defers for
4606 both kinds of timeout. A transport filter timeout is now identified in
4609 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4610 systems where "make" and "gmake" are different, calling "gmake" at top
4611 level broke things. I've arranged for the value of $(MAKE) to be passed
4612 from the Makefile to this script so that it can call the same version of
4616 A note about Exim versions 4.44 and 4.50
4617 ----------------------------------------
4619 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4620 changes of various kinds. As a consequence, a big documentation update was
4621 needed. This delayed the release for rather longer than seemed good, especially
4622 in the light of a couple of (minor) security issues. Therefore, the changes
4623 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4624 release. So 4.44 and 4.50 are in effect two different branches that both start
4627 I have left the 4.50 change log unchanged; it contains all the changes since
4628 4.43. The change log for 4.44 is below; many of its items are identical to
4629 those for 4.50. This seems to be the most sensible way to preserve the
4630 historical information.
4636 1. Minor wording change to the doc/README.SIEVE file.
4638 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4639 computation of the current number of files was incorrect.
4641 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4642 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4645 4. Give more explanation in the error message when the command for a transport
4646 filter fails to execute.
4648 5. There are several places where Exim runs a non-Exim command in a
4649 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4650 was being done only for the command run by the queryprogram router. It is
4651 now done for all such subprocesses. The other cases are: ${run, transport
4652 filters, and the commands run by the lmtp and pipe transports.
4654 6. Added CONFIGURE_GROUP build-time option.
4656 7. Some older OS have a limit of 256 on the maximum number of file
4657 descriptors. Exim was using setrlimit() to set 1000 as a large value
4658 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4659 systems. I've change it so that if it can't get 1000, it tries for 256.
4661 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4662 was an oversight, and furthermore, ever since the addition of extra
4663 controls (e.g. 4.43/32), the checks on when to allow different forms of
4664 "control" were broken. There should now be diagnostics for all cases when a
4665 control that does not make sense is encountered.
4667 9. Added the /retain_sender option to "control=submission".
4669 10. $recipients is now available in the predata ACL (oversight).
4671 11. Tidy the search cache before the fork to do a delivery from a message
4672 received from the command line. Otherwise the child will trigger a lookup
4673 failure and thereby defer the delivery if it tries to use (for example) a
4674 cached ldap connection that the parent has called unbind on.
4676 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4677 of $address_data from the recipient verification was clobbered by the
4678 sender verification.
4680 13. The value of address_data from a sender verification is now available in
4681 $sender_address_data in subsequent conditions in the ACL statement.
4683 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4685 15. Added a new option "connect=<time>" to callout options, to set a different
4688 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4689 was its contents. (It was OK if the option was not defined at all.)
4691 17. A "Completed" log line is now written for messages that are removed from
4692 the spool by the -Mrm option.
4694 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4695 information about exactly what failed.
4697 19. Added -dd to debug only the daemon process.
4699 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4700 handles timeouts, both on the server side and network timeouts. Renamed the
4701 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4703 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4704 It is now set to "smtps".
4706 22. $host_address is now set to the target address during the checking of
4707 ignore_target_hosts.
4709 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4710 being passed; this would have caused $sender_host_name to have been used if
4711 matching the list had actually called for a host name (not very likely,
4712 since this list is usually IP addresses). A host name is now passed as
4715 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4716 code that specifies a non-restarting handler (typically sigaction() in
4717 modern systems) in an attempt to fix a rare and obscure crash bug.
4719 25. Narrowed the window for a race in the daemon that could cause it to ignore
4720 SIGCHLD signals. This is not a major problem, because they are used only to
4721 wake it up if nothing else does.
4723 26. A malformed maildirsize file could cause Exim to calculate negative values
4724 for the mailbox size or file count. Odd effects could occur as a result.
4725 The maildirsize information is now recalculated if the size or filecount
4728 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4729 support for a long time. Removed HAVE_SYS_VFS_H.
4731 28. Installed the latest version of exipick from John Jetmore.
4733 29. In an address list, if the pattern was not a regular expression, an empty
4734 subject address (from a bounce message) matched only if the pattern was an
4735 empty string. Non-empty patterns were not even tested. This was the wrong
4736 because it is perfectly reasonable to use an empty address as part of a
4737 database query. An empty address is now tested by patterns that are
4738 lookups. However, all the other forms of pattern expect the subject to
4739 contain a local part and a domain, and therefore, for them, an empty
4740 address still always fails if the pattern is not itself empty.
4742 30. Exim went into a mad DNS loop when attempting to do a callout where the
4743 host was specified on an smtp transport, and looking it up yielded more
4744 than one IP address.
4746 31. Re-factored the code for checking spool and log partition space into a
4747 function that finds that data and another that does the check. The former
4748 is then used to implement four new variables: $spool_space, $log_space,
4749 $spool_inodes, and $log_inodes.
4751 32. The RFC2047 encoding function was originally intended for short strings
4752 such as real names; it was not keeping to the 75-character limit for
4753 encoded words that the RFC imposes. It now respects the limit, and
4754 generates multiple encoded words if necessary. To be on the safe side, I
4755 have increased the buffer size for the ${rfc2047: expansion operator from
4758 33. It is now permitted to omit both strings after an "if" condition; if the
4759 condition is true, the result is "true". As before, when the second string
4760 is omitted, a false condition yields an empty string. This makes it less
4761 cumbersome to write custom ACL and router conditions.
4763 34. Failure to deliver a bounce message always caused it to be frozen, even if
4764 there was an errors_to setting on the router. The errors_to setting is now
4767 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4768 canonical form (fully expanded) before being placed in
4769 $sender_host_address.
4771 36. The table in the code that translates DNS record types into text (T_A to
4772 "A" for instance) was missing entries for NS and CNAME. It is just possible
4773 that this could have caused confusion if both these types were looked up
4774 for the same domain, because the text type is used as part of Exim's
4775 per-process caching. But the chance of anyone hitting this buglet seems
4778 37. The dnsdb lookup has been extended in a number of ways.
4780 (1) There is a new type, "zns", which walks up the domain tree until it
4781 finds some nameserver records. It should be used with care.
4783 (2) There is a new type, "mxh", which is like "mx" except that it returns
4784 just the host names, not the priorities.
4786 (3) It is now possible to give a list of domains (or IP addresses) to be
4787 looked up. The behaviour when one of the lookups defers can be
4788 controlled by a keyword.
4790 (4) It is now possible to specify the separator character for use when
4791 multiple records are returned.
4793 38. The dnslists ACL condition has been extended: it is now possible to supply
4794 a list of IP addresses and/or domains to be looked up in a particular DNS
4797 39. Added log_selector=+queue_time_overall.
4799 40. When running the queue in the test harness, wait just a tad after forking a
4800 delivery process, to get repeatability of debugging output.
4802 41. Include certificate and key file names in error message when GnuTLS fails
4803 to set them up, because the GnuTLS error message doesn't include the name
4804 of the failing file when there is a problem reading it.
4806 42. Allow both -bf and -bF in the same test run.
4808 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4810 44. The "Exiscan patch" is now merged into the mainline Exim source.
4812 45. Sometimes the final signoff response after QUIT could fail to get
4813 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4814 < 0 before doing a fflush(). This bug looks as though it goes back to the
4815 introduction of TLS in release 3.20, but "sometimes" must have been rare
4816 because the tests only now provoked it.
4818 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4819 (this can affect the format of dates).
4821 47. exim_tidydb, when checking for the continued existence of a message for
4822 which it has found a message-specific retry record, was not finding
4823 messages that were in split spool directories. Consequently, it was
4824 deleting retry records that should have stayed in existence.
4826 48. Steve fixed some bugs in eximstats.
4828 49. The SPA authentication driver was not abandoning authentication and moving
4829 on to the next authenticator when an expansion was forced to fail,
4830 contradicting the general specification for all authenticators. Instead it
4831 was generating a temporary error. It now behaves as specified.
4833 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4834 (the order specifies the preference for clients). The order is now AES256,
4835 AES128, 3DES, ARCFOUR128.
4837 51. Small patch to Sieve code - explicitly set From: when generating an
4840 52. Exim crashed if a remote delivery caused a very long error message to be
4841 recorded - for instance if somebody sent an entire SpamAssassin report back
4842 as a large number of 550 error lines. This bug was coincidentally fixed by
4843 increasing the size of one of Exim's internal buffers (big_buffer) that
4844 happened as part of the Exiscan merge. However, to be on the safe side, I
4845 have made the code more robust (and fixed the comments that describe what
4848 53. Now that there can be additional text after "Completed" in log lines (if
4849 the queue_time_overall log selector is set), a one-byte patch to exigrep
4850 was needed to allow it to recognize "Completed" as not the last thing in
4853 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4854 patch that reportedly fixes this has been added. I am not expert enough to
4855 create a test for it. This is what the patch creator wrote:
4857 "I found a little strange behaviour of ldap code when working with
4858 Windows 2003 AD Domain, where users was placed in more than one
4859 Organization Units. When I tried to give exim partial DN, the exit code
4860 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4861 But simultaneously result of request was absolutely normal ldap result,
4862 so I produce this patch..."
4864 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4865 REFERENCE, so I have modified the code to exclude the patch when that macro
4868 55. Some experimental protocols are using DNS PTR records for new purposes. The
4869 keys for these records are domain names, not reversed IP addresses. The
4870 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4871 leaves it alone. Component reversal etc. now happens only for IP addresses.
4874 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4876 57. Double the size of the debug message buffer (to 2048) so that more of very
4877 long debug lines gets shown.
4879 58. The exicyclog utility now does better if the number of log files to keep
4880 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4882 59. Two changes related to the smtp_active_hostname option:
4884 (1) $smtp_active_hostname is now available as a variable.
4885 (2) The default for smtp_banner uses $smtp_active_hostname instead
4886 of $primary_hostname.
4888 60. The host_aton() function is supposed to be passed a string that is known
4889 to be a valid IP address. However, in the case of IPv6 addresses, it was
4890 not checking this. This is a hostage to fortune. Exim now panics and dies
4891 if the condition is not met. A case was found where this could be provoked
4892 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4893 components; fortuitously, this particular loophole had already been fixed
4894 by change 4.50/55 above.
4896 If there are any other similar loopholes, the new check in host_aton()
4897 itself should stop them being exploited. The report I received stated that
4898 data on the command line could provoke the exploit when Exim was running as
4899 exim, but did not say which command line option was involved. All I could
4900 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4901 running as the user.
4904 61. There was a buffer overflow vulnerability in the SPA authentication code
4905 (which came originally from the Samba project). I have added a test to the
4906 spa_base64_to_bits() function which I hope fixes it.
4909 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4910 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4912 63. The daemon start-up calls getloadavg() while still root for those OS that
4913 need the first call to be done as root, but it missed one case: when
4914 deliver_queue_load_max is set with deliver_drop_privilege. This is
4915 necessary for the benefit of the queue runner, because there is no re-exec
4916 when deliver_drop_privilege is set.
4918 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4919 This has been fixed.
4921 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4922 was in use, was not putting the data itself into the right store pool;
4923 consequently, it could be overwritten for a subsequent message in the same
4924 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4927 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4929 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4930 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4931 the function were treating the return as a boolean value, which happened to
4932 work because 0=false and not-0=true, but is not correct code.
4934 68. The host_aton() function was not handling scoped IPv6 addresses (those
4935 with, for example, "%eth0" on the end) correctly.
4937 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4938 negated items (that is, ~something) in unsigned ints. Some compilers
4939 apparently mutter when there is no cast.
4941 70. If an address verification called from an ACL failed, and did not produce a
4942 user-specific message (i.e. there was only a "system" message), nothing was
4943 put in $acl_verify_message. In this situation, it now puts the system
4946 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4947 session; change 4.31/43 added the unwanted input to the log line - except
4948 that it did not do this in the start of session case. It now does.
4950 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4951 This is wrong and can cause the other end to generate a synchronization
4952 error if it is another Exim or anything else that does the synchronization
4953 check. A QUIT command is no longer sent after a timeout.
4955 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4956 during host lookups.
4958 74. The defer_ok option of callout verification was not working if it was used
4959 when verifying addresses in header lines, that is, for this case:
4961 verify = header_sender/callout=defer_ok
4963 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4964 those file descriptors could be used for SMTP connections. If anything
4965 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4966 could be sent to the SMTP client, causing chaos. The daemon now opens
4967 stdin, stdout, and stderr to /dev/null when it puts itself into the
4970 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4971 log by default. The user can override this with suitable Perl magic.
4973 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4974 the log message when discard triggers, was not working for the DATA ACL or
4975 for the non-SMTP ACL.
4977 78. Error message wording change in sieve.c.
4979 79. If smtp_accept_max_per_host was set, the number of connections could be
4980 restricted to fewer than expected, because the daemon was trying to set up
4981 a new connection before checking whether the processes handling previous
4982 connections had finished. The check for completed processes is now done
4983 earlier. On busy systems, this bug wouldn't be noticed because something
4984 else would have woken the daemon, and it would have reaped the completed
4987 80. If a message was submitted locally by a user whose login name contained one
4988 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4989 It caused a spool format error. I have fixed the spool reading code. A
4990 related problem was that the "from" clause in the Received: line became
4991 illegal because of the space(s). It is now covered by ${quote_local_part.
4993 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4996 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4998 83. Patch from Sieve maintainer to fix unterminated string problem in
4999 "vacation" handling.
5001 84. Some minor changes to the Linux configuration files to help with other
5002 OS variants using glibc.
5004 85. One more patch for Sieve to update vacation handling to latest spec.
5007 ----------------------------------------------------
5008 See the note above about the 4.44 and 4.50 releases.
5009 ----------------------------------------------------
5015 1. Change 4.43/35 introduced a bug that caused file counts to be
5016 incorrectly computed when quota_filecount was set in an appendfile
5019 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5020 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5023 3. Give more explanation in the error message when the command for a transport
5024 filter fails to execute.
5026 4. There are several places where Exim runs a non-Exim command in a
5027 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5028 was being done only for the command run by the queryprogram router. It is
5029 now done for all such subprocesses. The other cases are: ${run, transport
5030 filters, and the commands run by the lmtp and pipe transports.
5032 5. Some older OS have a limit of 256 on the maximum number of file
5033 descriptors. Exim was using setrlimit() to set 1000 as a large value
5034 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5035 systems. I've change it so that if it can't get 1000, it tries for 256.
5037 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5038 was an oversight, and furthermore, ever since the addition of extra
5039 controls (e.g. 4.43/32), the checks on when to allow different forms of
5040 "control" were broken. There should now be diagnostics for all cases when a
5041 control that does not make sense is encountered.
5043 7. $recipients is now available in the predata ACL (oversight).
5045 8. Tidy the search cache before the fork to do a delivery from a message
5046 received from the command line. Otherwise the child will trigger a lookup
5047 failure and thereby defer the delivery if it tries to use (for example) a
5048 cached ldap connection that the parent has called unbind on.
5050 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5051 of $address_data from the recipient verification was clobbered by the
5052 sender verification.
5054 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5055 was its contents. (It was OK if the option was not defined at all.)
5057 11. A "Completed" log line is now written for messages that are removed from
5058 the spool by the -Mrm option.
5060 12. $host_address is now set to the target address during the checking of
5061 ignore_target_hosts.
5063 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5064 being passed; this would have caused $sender_host_name to have been used if
5065 matching the list had actually called for a host name (not very likely,
5066 since this list is usually IP addresses). A host name is now passed as
5069 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5070 code that specifies a non-restarting handler (typically sigaction() in
5071 modern systems) in an attempt to fix a rare and obscure crash bug.
5073 15. Narrowed the window for a race in the daemon that could cause it to ignore
5074 SIGCHLD signals. This is not a major problem, because they are used only to
5075 wake it up if nothing else does.
5077 16. A malformed maildirsize file could cause Exim to calculate negative values
5078 for the mailbox size or file count. Odd effects could occur as a result.
5079 The maildirsize information is now recalculated if the size or filecount
5082 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5083 support for a long time. Removed HAVE_SYS_VFS_H.
5085 18. Updated exipick to current release from John Jetmore.
5087 19. Allow an empty sender to be matched against a lookup in an address list.
5088 Previously the only cases considered were a regular expression, or an
5091 20. Exim went into a mad DNS lookup loop when doing a callout where the
5092 host was specified on the transport, if the DNS lookup yielded more than
5095 21. The RFC2047 encoding function was originally intended for short strings
5096 such as real names; it was not keeping to the 75-character limit for
5097 encoded words that the RFC imposes. It now respects the limit, and
5098 generates multiple encoded words if necessary. To be on the safe side, I
5099 have increased the buffer size for the ${rfc2047: expansion operator from
5102 22. Failure to deliver a bounce message always caused it to be frozen, even if
5103 there was an errors_to setting on the router. The errors_to setting is now
5106 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5107 canonical form (fully expanded) before being placed in
5108 $sender_host_address.
5110 24. Updated eximstats to version 1.33
5112 25. Include certificate and key file names in error message when GnuTLS fails
5113 to set them up, because the GnuTLS error message doesn't include the name
5114 of the failing file when there is a problem reading it.
5116 26. Expand error message when OpenSSL has problems setting up cert/key files.
5119 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5120 (this can affect the format of dates).
5122 28. exim_tidydb, when checking for the continued existence of a message for
5123 which it has found a message-specific retry record, was not finding
5124 messages that were in split spool directories. Consequently, it was
5125 deleting retry records that should have stayed in existence.
5127 29. eximstats updated to version 1.35
5128 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5129 1.35 - bugfix such that pie charts by volume are generated correctly
5131 30. The SPA authentication driver was not abandoning authentication and moving
5132 on to the next authenticator when an expansion was forced to fail,
5133 contradicting the general specification for all authenticators. Instead it
5134 was generating a temporary error. It now behaves as specified.
5136 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5137 (the order specifies the preference for clients). The order is now AES256,
5138 AES128, 3DES, ARCFOUR128.
5140 31. Small patch to Sieve code - explicitly set From: when generating an
5143 32. Exim crashed if a remote delivery caused a very long error message to be
5144 recorded - for instance if somebody sent an entire SpamAssassin report back
5145 as a large number of 550 error lines. This bug was coincidentally fixed by
5146 increasing the size of one of Exim's internal buffers (big_buffer) that
5147 happened as part of the Exiscan merge. However, to be on the safe side, I
5148 have made the code more robust (and fixed the comments that describe what
5151 33. Some experimental protocols are using DNS PTR records for new purposes. The
5152 keys for these records are domain names, not reversed IP addresses. The
5153 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5154 leaves it alone. Component reversal etc. now happens only for IP addresses.
5157 34. The host_aton() function is supposed to be passed a string that is known
5158 to be a valid IP address. However, in the case of IPv6 addresses, it was
5159 not checking this. This is a hostage to fortune. Exim now panics and dies
5160 if the condition is not met. A case was found where this could be provoked
5161 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5162 components; fortuitously, this particular loophole had already been fixed
5163 by change 4.50/55 or 4.44/33 above.
5165 If there are any other similar loopholes, the new check in host_aton()
5166 itself should stop them being exploited. The report I received stated that
5167 data on the command line could provoke the exploit when Exim was running as
5168 exim, but did not say which command line option was involved. All I could
5169 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5170 running as the user.
5173 35. There was a buffer overflow vulnerability in the SPA authentication code
5174 (which came originally from the Samba project). I have added a test to the
5175 spa_base64_to_bits() function which I hope fixes it.
5178 36. The daemon start-up calls getloadavg() while still root for those OS that
5179 need the first call to be done as root, but it missed one case: when
5180 deliver_queue_load_max is set with deliver_drop_privilege. This is
5181 necessary for the benefit of the queue runner, because there is no re-exec
5182 when deliver_drop_privilege is set.
5184 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5185 was in use, was not putting the data itself into the right store pool;
5186 consequently, it could be overwritten for a subsequent message in the same
5187 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5190 38. Sometimes the final signoff response after QUIT could fail to get
5191 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5192 < 0 before doing a fflush(). This bug looks as though it goes back to the
5193 introduction of TLS in release 3.20, but "sometimes" must have been rare
5194 because the tests only now provoked it.
5200 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5201 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5202 Now it returns an integer. A number of calls were still expecting a T/F
5203 return. Fortuitously, in all cases, the tests worked in OK situations,
5204 which is the norm. However, things would have gone wrong on any write
5205 failures on the smtp file descriptor. This function is used when sending
5206 messages over SMTP and also when doing verify callouts.
5208 2. When Exim is called to do synchronous delivery of a locally submitted
5209 message (the -odf or -odi options), it no longer closes stderr before doing
5212 3. Implemented the mua_wrapper option.
5214 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5216 5. Implemented the functions header_remove(), header_testname(),
5217 header_add_at_position(), and receive_remove_recipient(), and exported them
5220 6. If an ACL "warn" statement specified the addition of headers, Exim already
5221 inserted X-ACL-Warn: at the start if there was no header name. However, it
5222 was not making this test for the second and subsequent header lines if
5223 there were newlines in the string. This meant that an invalid header could
5224 be inserted if Exim was badly configured.
5226 7. Allow an ACL "warn" statement to add header lines at the start or after all
5227 the Received: headers, as well as at the end.
5229 8. Added the rcpt_4xx retry error code.
5231 9. Added postmaster_mailfrom=xxx to callout verification option.
5233 10. Added mailfrom=xxxx to the callout verification option, for verify=
5236 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5237 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5239 12. Inserted some casts to stop certain compilers warning when using pointer
5240 differences as field lengths or precisions in printf-type calls (mostly
5241 affecting debugging statements).
5243 13. Added optional readline() support for -be (dynamically loaded).
5245 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5246 same clock tick as a message's arrival, so that its received time was the
5247 same as the "first fail" time on the retry record, and that message
5248 remained on the queue past the ultimate address timeout, every queue runner
5249 would try a delivery (because it was past the ultimate address timeout) but
5250 after another failure, the ultimate address timeout, which should have then
5251 bounced the address, did not kick in. This was a "< instead of <=" error;
5252 in most cases the first failure would have been in the next clock tick
5253 after the received time, and all would be well.
5255 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5256 being recognized when the domain list was tested by the match_domain
5257 condition in an expansion string.
5259 16. Added the ${str2b64: operator.
5261 17. Exim was always calling setrlimit() to set a large limit for the number of
5262 processes, without checking whether the existing limit was already
5263 adequate. (It did check for the limit on file descriptors.) Furthermore,
5264 errors from getrlimit() and setrlimit() were being ignored. Now they are
5265 logged to the main and panic logs, but Exim does carry on, to try to do its
5266 job under whatever limits there are.
5268 18. Imported PCRE 5.0.
5270 19. Trivial typo in log message " temporarily refused connection" (the leading
5273 20. If the log selector return_path_on_delivery was set and an address was
5274 redirected to /dev/null, the delivery process crashed because it assumed
5275 that a return path would always be set for a "successful" delivery. In this
5276 case, the whole delivery is bypassed as an optimization, and therefore no
5279 21. Internal re-arrangement: the function for sending a challenge and reading
5280 a response while authentication was assuming a zero-terminated challenge
5281 string. It's now changed to take a pointer and a length, to allow for
5282 binary data in such strings.
5284 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5286 23. Exim was not respecting finduser_retries when seeking the login of the
5287 uid under which it was called; it was always trying 10 times. (The default
5288 setting of finduser_retries is zero.) Also, it was sleeping after the final
5289 failure, which is pointless.
5291 24. Implemented tls_on_connect_ports.
5293 25. Implemented acl_smtp_predata.
5295 26. If the domain in control=submission is set empty, Exim assumes that the
5296 authenticated id is a complete email address when it generates From: or
5297 Sender: header lines.
5299 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5300 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5301 chown and chgrp in /bin and hostname in /usr/bin.
5303 28. Exim was keeping the "process log" file open after each use, just as it
5304 does for the main log. This opens the possibility of it remaining open for
5305 long periods when the USR1 signal hits a daemon. Occasional processlog
5306 errors were reported, that could have been caused by this. Anyway, it seems
5307 much more sensible not to leave this file open at all, so that is what now
5310 29. The long-running daemon process does not normally write to the log once it
5311 has entered its main loop, and it closes the log before doing so. This is
5312 so that log files can straightforwardly be renamed and moved. However,
5313 there are a couple of unusual error situations where the daemon does write
5314 log entries, and I had neglected to close the log afterwards.
5316 30. The text of an SMTP error response that was received during a remote
5317 delivery was being truncated at 512 bytes. This is too short for some of
5318 the long messages that one sometimes sees. I've increased the limit to
5321 31. It is now possible to make retry rules that apply only when a message has a
5322 specific sender, in particular, an empty sender.
5324 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5325 it possible to be selective about when SMTP synchronization is enforced.
5327 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5329 32. Implemented hosts_connection_nolog.
5331 33. Added an ACL for QUIT.
5333 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5336 35. Added mailbox_size and mailbox_filecount to appendfile.
5338 36. Added control = no_multiline_responses to ACLs.
5340 37. There was a bug in the logic of the code that waits for the clock to tick
5341 in the case where the clock went backwards by a substantial amount such
5342 that the microsecond fraction of "now" was more than the microsecond
5343 fraction of "then" (but the whole seconds number was less).
5345 38. Added support for the libradius Radius client library this is found on
5346 FreeBSD (previously only the radiusclient library was supported).
5352 1. When certain lookups returned multiple values in the form name=value, the
5353 quoting of the values was not always being done properly. Specifically:
5354 (a) If the value started with a double quote, but contained no whitespace,
5356 (b) If the value contained whitespace other than a space character (i.e.
5357 tabs or newlines or carriage returns) it was not quoted.
5358 This fix has been applied to the mysql and pgsql lookups by writing a
5359 separate quoting function and calling it from the lookup code. The fix
5360 should probably also be applied to nisplus, ibase and oracle lookups, but
5361 since I cannot test any of those, I have not disturbed their existing code.
5363 2. A hit in the callout cache for a specific address caused a log line with no
5364 reason for rejecting RCPT. Now it says "Previous (cached) callout
5365 verification failure".
5367 3. There was an off-by-one bug in the queryprogram router. An over-long
5368 return line was truncated at 256 instead of 255 characters, thereby
5369 overflowing its buffer with the terminating zero. As well as fixing this, I
5370 have increased the buffer size to 1024 (and made a note to document this).
5372 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5373 when Exim is waiting for an SMTP response from a remote server, Exim
5374 restarts its select() call on the socket, thereby resetting its timeout.
5375 This is not a problem when such interrupts are rare. Somebody set up a cron
5376 job to run exiwhat every 2 minutes, which is less than the normal select()
5377 timeout (5 or 10 minutes). This meant that the select() timeout never
5378 kicked in because it was always reset. I have fixed this by comparing the
5379 time when an interrupt arrives with the time at the start of the first call
5380 to select(). If more time than the timeout has elapsed, the interrupt is
5381 treated as a timeout.
5383 5. Some internal re-factoring in preparation for the addition of Sieve
5384 extensions (by MH). In particular, the "personal" test is moved to a
5385 separate function, and given an option for scanning Cc: and Bcc: (which is
5386 not set for Exim filters).
5388 6. When Exim created an email address using the login of the caller as the
5389 local part (e.g. when creating a From: or Sender: header line), it was not
5390 quoting the local part when it contained special characters such as @.
5392 7. Installed new OpenBSD configuration files.
5394 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5395 try to make them clearer.
5397 9. Callout options, other than the timeout value, were being ignored when
5398 verifying sender addresses in header lines. For example, when using
5400 verify = header_sender/callout=no_cache
5402 the cache was (incorrectly) being used.
5404 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5405 only the Cygwin environment.
5407 11. When return_path_on_delivery was set as a log selector, if different remote
5408 addresses in the same message used different return paths and parallel
5409 remote delivery occurred, the wrong values would sometimes be logged.
5410 (Whenever a remote delivery process finished, the return path value from
5411 the most recently started remote delivery process was logged.)
5413 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5414 lines when AUTH and/or TLS are in use. This is the "received protocol"
5415 field. Exim used to use "asmtp" for authenticated SMTP, without any
5416 indication (in the protocol name) for TLS use. Now it follows the RFC and
5417 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5418 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5419 names appear in log lines as well as in Received: header lines.
5421 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5422 extensions, and comparison tests, and to fix some bugs.
5424 14. Changes to the "personal" filter test:
5426 (1) The test was buggy in that it was just doing the equivalent of
5427 "contains" tests on header lines. For example, if a user's address was
5428 anne@some.where, the "personal" test would incorrectly be true for
5430 To: susanne@some.where
5432 This test is now done by extracting each address from the header in turn,
5433 and checking the entire address. Other tests that are part of "personal"
5434 are now done using regular expressions (for example, to check local parts
5435 of addresses in From: header lines).
5437 (2) The list of non-personal local parts in From: addresses has been
5438 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5439 taken from the Sieve specification recommendations.
5441 (3) If the message contains any header line starting with "List-" it is
5442 treated as non-personal.
5444 (4) The test for "circular" in the Subject: header line has been removed
5445 because it now seems ill-conceived.
5447 15. Minor typos in src/EDITME comments corrected.
5449 16. Installed latest exipick from John Jetmore.
5451 17. If headers_add on a router specified a text string that was too long for
5452 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5453 of string_sprintf() is now avoided.
5455 18. $message_body_size was not set (it was always zero) when running the DATA
5456 ACL and the local_scan() function.
5458 19. For the "mail" command in an Exim filter, no default was being set for
5459 the once_repeat time, causing a random time value to be used if "once" was
5460 specified. (If the value happened to be <= 0, no repeat happened.) The
5461 default is now 0s, meaning "never repeat". The "vacation" command was OK
5462 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5463 (I found it when inspecting the code).
5465 20. There is now an overall timeout for performing a callout verification. It
5466 defaults to 4 times the callout timeout, which applies to individual SMTP
5467 commands during the callout. The overall timeout applies when there is more
5468 than one host that can be tried. The timeout is checked before trying the
5469 next host. This prevents very long delays if there are a large number of
5470 hosts and all are timing out (e.g. when the network connections are timing
5471 out). The value of the overall timeout can be changed by specifying an
5472 additional sub-option for "callout", called "maxwait". For example:
5474 verify = sender/callout=5s,maxwait=20s
5476 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5477 to the end before writing, but this should make it even safer).
5479 22. Exim was forgetting that it had advertised PIPELINING for the second and
5480 subsequent messages on an SMTP connection. It was also not resetting its
5481 memory on STARTTLS and an internal HELO.
5483 23. When Exim logs an SMTP synchronization error within a session, it now
5484 records whether PIPELINING has been advertised or not.
5486 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5487 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5488 rather than long int.
5490 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5492 26. Added the never_mail option to autoreply.
5498 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5499 crash if the getsockname() call failed; this can happen if a connection is
5500 closed very soon after it is established. The problem was simply in the
5501 order in which certain operations were done, causing Exim to try to write
5502 to the SMTP stream before it had set up the file descriptor. The bug has
5503 been fixed by making things happen in the correct order.
5509 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5510 before the connection was closed, thus losing the rejection response.
5512 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5513 some early Solaris releases, but causes trouble in current releases where
5514 socklen_t is defined.
5516 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5519 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5522 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5523 that releasing the top of it at the end releases what was used for sub-
5524 expansions (unless the block got too big). However, discard this block if
5525 the first thing is a variable or header, so that we can use its block when
5526 it is dynamic (useful for very large $message_headers, for example).
5528 6. Lookups now cache *every* query, not just the most recent. A new, separate
5529 store pool is used for this. It can be recovered when all lookup caches are
5530 flushed. Lookups now release memory at the end of their result strings.
5531 This has involved some general refactoring of the lookup sources.
5533 7. Some code has been added to the store_xxx() functions to reduce the amount
5534 of flapping under certain conditions.
5536 8. log_incoming_interface used to affect only the <= reception log lines. Now
5537 it causes the local interface and port to be added to several more SMTP log
5538 lines, for example "SMTP connection from", and rejection lines.
5540 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5542 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5544 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5545 could be overwritten at the end of the current message (or the start of a
5546 new message if it was set in a HELO ACL). The value is now preserved for
5547 the duration of the SMTP connection.
5549 12. If a transport had a headers_rewrite setting, and a matching header line
5550 contained an unqualified address, that address was qualified, even if it
5551 did not match any rewriting rules. The underlying bug was that the values
5552 of the flags that permit the existence of unqualified sender and recipient
5553 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5554 non-local messages, and by -bnq for local messages) were not being
5555 preserved with the message after it was received.
5557 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5558 "next input=" as part of the text comprising the host identity instead of
5559 the correct text. The code was using the same buffer for two different
5560 strings. However, depending on which order the printing function evaluated
5561 its arguments, the bug did not always show up. Under Linux, for example, my
5562 test suite worked just fine.
5564 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5565 doesn't work with some older versions of Perl. It has been changed to "my",
5566 which in any case is probably the better facility to use.
5568 15. A really picky compiler found some instances of statements for creating
5569 error messages that either had too many or two few arguments for the format
5572 16. The size of the buffer for calls to the DNS resolver has been increased
5573 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5574 for addresses that have a lot of PTR records. This alleviates a problem; it
5575 does not fully solve it.
5577 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5578 buffer now truncates the list and logs the incident, which is the same
5579 action as happens when Exim is looking up a host name and its aliases.
5580 Previously in this situation something unpredictable would happen;
5581 sometimes it was "internal error: store_reset failed".
5583 18. If a server dropped the connection unexpectedly when an Exim client was
5584 using GnuTLS and trying to read a response, the client delivery process
5585 crashed while trying to generate an error log message.
5587 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5588 string, for example:
5590 warn message = H1: something\nH2: something
5592 the text was added as a single header line from Exim's point of view
5593 though it ended up OK in the delivered message. However, searching for the
5594 second and subsequent header lines using $h_h2: did not work. This has been
5595 fixed. Similarly, if a system filter added multiple headers in this way,
5596 the routers could not see them.
5598 20. Expanded the error message when iplsearch is called with an invalid key to
5599 suggest using net-iplsearch in a host list.
5601 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5602 ACLs are no longer actually imposed (and a message to that effect is
5605 22. If a "gecos" field in a passwd entry contained escaped characters, in
5606 particular, if it contained a \" sequence, Exim got it wrong when building
5607 a From: or a Sender: header from that name. A second bug also caused
5608 incorrect handling when an unquoted " was present following a character
5609 that needed quoting.
5611 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5612 was not being matched caselessly.
5614 24. Arranged for all hyphens in the exim.8 source to be escaped with
5617 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5618 properly. Recipient callout cache records were still being keyed to include
5619 the sender, even when use_sender was set false. This led to far more
5620 callouts that were necessary. The sender is no longer included in the key
5621 when use_sender is false.
5623 26. Added "control = submission" modifier to ACLs.
5625 27. Added the ${base62d: operator to decode base 62 numbers.
5627 28. dnsdb lookups can now access SRV records.
5629 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5630 the configuration file.
5632 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5633 (-v) mode. This makes the output for a verbose queue run more intelligible.
5635 31. Added a use_postmaster feature to recipient callouts.
5637 32. Added the $body_zerocount variable, containing the number of binary zero
5638 bytes in the message body.
5640 33. The time of last modification of the "new" subdirectory is now used as the
5641 "mailbox time last read" when there is a quota error for a maildir
5644 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5646 35. Added +ignore_unknown as a special item in host lists.
5648 36. Code for decoding IPv6 addresses in host lists is now included, even if
5649 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5650 address was recognized as an IP address, but was then not correctly decoded
5651 into binary, causing unexpected and incorrect effects when compared with
5658 1. Very minor rewording of debugging text in manualroute to say "list of
5659 hosts" instead of "hostlist".
5661 2. If verify=header_syntax was set, and a header line with an unqualified
5662 address (no domain) and a large number of spaces between the end of the
5663 name and the colon was received, the reception process suffered a buffer
5664 overflow, and (when I tested it) crashed. This was caused by some obsolete
5665 code that should have been removed. The fix is to remove it!
5667 3. When running in the test harness, delay a bit after writing a bounce
5668 message to get a bit more predictability in the log output.
5670 4. Added a call to search_tidyup() just before forking a reception process. In
5671 theory, someone could use a lookup in the expansion of smtp_accept_max_
5672 per_host which, without the tidyup, could leave open a database connection.
5674 5. Added the variables $recipient_data and $sender_data which get set from a
5675 lookup success in an ACL "recipients" or "senders" condition, or a router
5676 "senders" option, similar to $domain_data and $local_part_data.
5678 6. Moved the writing of debug_print from before to after the "senders" test
5681 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5682 problems for message scanning, either using a data ACL, or using
5683 local_scan() because the Received: header was not generated till after they
5684 were called (in order to set the time as the time of reception completion).
5685 I have revised the way this works. The header is now generated after the
5686 body is received, but before the ACL or local_scan() are called. After they
5687 are run, the timestamp in the header is updated.
5693 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5694 before starting a queue runner without re-exec. This happened only when
5695 deliver_drop_privilege was set or when the Exim user was set to root. The
5696 effect of the bug was that timeouts during subsequent deliveries caused
5697 crashes instead of being properly handled. The handler is now left at its
5698 default (and expected) setting.
5700 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5701 message, again when deliver_drop_privilege is set or Exim is run as root.
5702 The bug described in (1) was not present in this case, but the tidying up
5703 of the other signals was missing. I have made the two cases consistent.
5705 3. The ignore_target_hosts setting on a manualroute router was being ignored
5706 for hosts that were looked up using the /MX notation.
5708 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5711 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5712 operated on the sender address. After changing the $sender_address to <>
5713 for the sender address verify, Exim was re-instated it as the original
5714 (before rewriting) address, but remembering that it had rewritten it, so it
5715 wasn't rewriting it again. This bug also had the effect of breaking the
5716 sender address verification caching when the sender address was rewritten.
5718 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5719 This has been changed so that if the ip literal address matches
5720 ignore_target_hosts, the router declines.
5722 7. Added expansion conditions match_domain, match_address, and match_local_
5723 part (NOT match_host).
5725 8. The placeholder for the Received: header didn't have a length field set.
5727 9. Added code to Exim itself and to exim_lock to test for a specific race
5728 condition that could lead to file corruption when using MBX delivery. The
5729 issue is with the lockfile that is created in /tmp. If this file is removed
5730 after a process has opened it but before that process has acquired a lock,
5731 there is the potential for a second process to recreate the file and also
5732 acquire a lock. This could lead to two Exim processes writing to the file
5733 at the same time. The added code performs the same test as UW imapd; it
5734 checks after acquiring the lock that its file descriptor still refers to
5735 the same named file.
5737 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5738 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5741 11. Added the smtp_active_hostname option. If used, this will typically be made
5742 to depend on the incoming interface address. Because $interface_address is
5743 not set up until the daemon has forked a reception process, error responses
5744 that can happen earlier (such as "too many connections") no longer contain
5747 12. If an expansion in a condition on a "warn" statement fails because a lookup
5748 defers, the "warn" statement is abandoned, and the next ACL statement is
5749 processed. Previously this caused the whole ACL to be aborted.
5751 13. Added the iplsearch lookup type.
5753 14. Added ident_timeout as a log selector.
5755 15. Added tls_certificate_verified as a log selector.
5757 16. Added a global option tls_require_ciphers (compare the smtp transport
5758 option of the same name). This controls incoming TLS connections.
5760 17. I finally figured out how to make tls_require_ciphers do a similar thing
5761 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5762 before starting the TLS session.
5764 18. Tabs are now shown as \t in -bP output.
5766 19. If the log selector return_path_on_delivery was set, Exim crashed when
5767 bouncing a message because it had too many Received: header lines.
5769 20. If two routers both had headers_remove settings, and the first one included
5770 a superfluous trailing colon, the final name in the first list and the
5771 first name in the second list were incorrectly joined into one item (with a
5772 colon in the middle).
5778 1. Added -C and -D options to the exinext utility, mainly to make it easier
5779 to include in the automated testing, but these could be helpful when
5780 multiple configurations are in use.
5782 2. The exinext utility was not formatting the output nicely when there was
5783 an alternate port involved in the retry record key, nor when there was a
5784 message id as well (for retries that were specific to a specific message
5785 and a specific host). It was also confused by IPv6 addresses, because of
5786 the additional colons they contain. I have fixed the IPv4 problem, and
5787 patched it up to do a reasonable job for IPv6.
5789 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5790 delivery, the log line now contains "pipelined" if PIPELINING was used.
5792 4. An SMTP transport process used to panic and die if the bind() call to set
5793 an explicit outgoing interface failed. This has been changed; it is now
5794 treated in the same way as a connect() failure.
5796 5. A reference to $sender_host_name in the part of a conditional expansion
5797 that was being skipped was still causing a DNS lookup. This no longer
5800 6. The def: expansion condition was not recognizing references to header lines
5801 that used bh_ and bheader_.
5803 7. Added the _cache feature to named lists.
5805 8. The code for checking quota_filecount in the appendfile transport was
5806 allowing one more file than it should have been.
5808 9. For compatibility with Sendmail, the command line option
5816 and sets the incoming protocol and host name (for trusted callers). The
5817 host name and its colon can be omitted when only the protocol is to be set.
5818 Note the Exim already has two private options, -pd and -ps, that refer to
5819 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5820 "s", but I don't think that's a major issue.
5822 10. A number of refactoring changes to the code, none of which should affect
5825 (a) The number of logging options was getting close to filling up the
5826 32-bit word that was used as a bit map. I have split them into two classes:
5827 those that are passed in the argument to log_write(), and those that are
5828 only ever tested independently outside of that function. These are now in
5829 separate 32-bit words, so there is plenty of room for expansion again.
5830 There is no change in the user interface or the logging behaviour.
5832 (b) When building, for example, log lines, the code previously used a
5833 macro that called string_cat() twice, in order to add two strings. This is
5834 not really sufficiently general. Furthermore, there was one instance where
5835 it was actually wrong because one of the argument was used twice, and in
5836 one call a function was used. (As it happened, calling the function twice
5837 did not affect the overall behaviour.) The macro has been replaced by a
5838 function that can join an arbitrary number of extra strings onto a growing
5841 (c) The code for expansion conditions now uses a table and a binary chop
5842 instead of a serial search (which was left over from when there were very
5843 few conditions). Also, it now recognizes conditions like "pam" even when
5844 the relevant support is not compiled in: a suitably worded error message is
5845 given if an attempt is made to use such a condition.
5847 11. Added ${time_interval:xxxxx}.
5849 12. A bug was causing one of the ddress fields not to be passed back correctly
5850 from remote delivery subprocesses. The field in question was not being
5851 subsequently used, so this caused to problems in practice.
5853 13. Added new log selectors queue_time and deliver_time.
5855 14. Might have fixed a bug in maildirsizefile handling that threw up
5856 "unexpected character" debug warnings, and recalculated the data
5857 unnecessarily. In any case, I expanded the warning message to give more
5860 15. Added the message "Restricted characters in address" to the statements in
5861 the default ACL that block characters like @ and % in local parts.
5863 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5864 Three changes have been made:
5866 (a) There was a serious bug; a negative response to MAIL caused the whole
5867 recipient domain to be cached as invalid, thereby blocking all messages
5868 to all local parts at the same domain, from all senders. This bug has
5869 been fixed. The domain is no longer cached after a negative response to
5870 MAIL if the sender used is not empty.
5872 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5875 (c) A new callout option, "use_sender" has been added for people who want
5876 the modified behaviour.
5882 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5885 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5886 indeed breaks things for older releases.
5888 3. Added additional logging to the case where there is a problem reading data
5889 from a filter that is running in a subprocess using a pipe, in order to
5890 try to track down a specific problem.
5892 4. Testing facility fudge: when running in the test harness and attempting
5893 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5894 getting "No route to host". Convert this to a timeout.
5896 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5899 6. Some OS don't have socklen_t but use size_t instead. This affects the
5900 fifth argument of getsockopt() amongst other things. This is now
5901 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5902 can be set for individual OS. I have set it for SunOS5, OSF1, and
5903 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5904 some earlier ones do not.
5906 7. Change 4.30/15 was not doing the test caselessly.
5908 8. The standard form for an IPv6 address literal was being rejected by address
5909 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5910 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5911 this, as well as the form without the "IPv6" on the front (but only when
5912 address literals are enabled, of course).
5914 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5916 10. Exim crashed if a message with an empty sender address specified by -f
5917 encountered a router with an errors_to setting. This could be provoked only
5918 by a command such as
5922 where an empty string was supplied; "<>" did not hit this bug.
5924 11. Installed PCRE release 4.5.
5926 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5927 remained set. It is now erased.
5929 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5930 times from message ids (which are base 36 rather than the normal 62).
5932 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5933 were being counted as actual protocol errors, and logged if the log
5934 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5935 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5936 and DATA following a set of rejected RCPTs do not count as protocol errors.
5937 In other words, Exim assumes they were pipelined, though this may not
5938 actually be the case. Of course, in all cases the client gets an
5939 appropriate error code.
5941 15. If a lookup fails in an ACL condition, a message about the failure may
5942 be available; it is used if testing the ACL cannot continue, because most
5943 such messages specify what the cause of the deferral is. However, some
5944 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5945 that caused an old message to be retained and used if a later statement
5946 caused a defer, replacing the real cause of the deferral.
5948 16. If an IP address had so many PTR records that the DNS lookup buffer
5949 was not large enough to hold them, Exim could crash while trying to process
5950 the truncated data. It now detects and logs this case.
5952 17. Further to 4.21/58, another change has been made: if (and only if) the
5953 first line of a message (the first header line) ends with CRLF, a bare LF
5954 in a subsequent header line has a space inserted after it, so as not to
5955 terminate the header.
5957 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5958 unnecessarily, used atoi() instead of strtol(), and didn't check the
5959 termination when getting file sizes from file names by regex.
5961 19. Completely re-implemented the support for maildirsize files, in the light
5962 of a number of problems with the previous contributed implementation
5963 (4.30/29). In particular:
5965 . If the quota is zero, the maildirsize file is maintained, but no quota is
5968 . If the maildir directory does not exist, it is created before any attempt
5969 to write a maildirsize file.
5971 . The quota value in the file is just a cache; if the quota is changed in
5972 the transport, the new value overrides.
5974 . A regular expression is available for excluding directories from the
5977 20. The autoreply transport checks the characters in options that define the
5978 message's headers; it allows continued headers, but it was checking with
5979 isspace() after an embedded newline instead of explicitly looking for a
5982 21. If all the "regular" hosts to which an address was routed had passed their
5983 expiry times, and had not reached their retry times, the address was
5984 bounced, even if fallback hosts were defined. Now Exim should go on to try
5987 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5988 equivalent code in the SMTP transport. Some hosts send humungous responses
5989 to HELO/EHLO, more than 1024 it seems.
5991 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5992 gives compiler warnings in some environments. I've now done it "properly",
5995 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5996 (because of problems with the built-in one) was declared to return uschar *
5997 instead of char *, causing compiler failure.
5999 25. Fixed a file descriptor leak when processing alias/forward files.
6001 26. Fixed a minor format string issue in dbfn.c.
6003 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6005 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6006 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6009 29. When a sender address is verified, it is cached, to save repeating the test
6010 when there is more than one recipient in a message. However, when the
6011 verification involves a callout, it is possible for different callout
6012 options to be set for different recipients. It is too complicated to keep
6013 track of this in the cache, so now Exim always runs a verification when a
6014 callout is required, relying on the callout cache for the optimization.
6015 The overhead is duplication of the address routing, but this should not be
6018 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6019 to be verified with callout=postmaster, and the main callout worked but the
6020 postmaster check failed, the verification correctly failed. However, if a
6021 subsequent RCPT command asked for sender verification *without* the
6022 postmaster check, incorrect caching caused this verification also to fail,
6025 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6026 it was not caching the DNS options (qualify_single, search_parents) that
6027 were used when the lookup failed. A subsequent lookup with different
6028 options therefore always gave the same answer, though there were cases
6029 where it should not have. (Example: a "domains = !$mx_any" option on a
6030 dnslookup router: the "domains" option is always processed without any
6031 widening, but the router might have qualify_single set.) Now Exim uses the
6032 cached value only when the same options are set.
6034 32. Added John Jetmore's "exipick" utility to the distribution.
6036 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6037 than a timeout (e.g. a certificate is required, and is not provided), an
6038 Exim server now closes the connection immediately. Previously it waited for
6039 the client to close - but if the client is SSL, it seems that they each
6040 wait for each other, leading to a delay before one of them times out.
6042 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6043 maintained 0.8.x compatibility because I don't think many are using it, and
6044 it is clearly obsolete.
6046 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6049 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6050 client certificate was expired. A simple patch fixes this, though I don't
6051 understand the full logic of why the verify callback is called multiple
6054 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6055 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6056 which causes problems with some clients (such as the Certicom SSL Plus
6057 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6058 disables the coutermeasure allowing Eudora to connect."
6060 38. Exim was not checking that a write() to a log file succeeded. This could
6061 lead to Bad Things if a log got too big, in particular if it hit a file
6062 size limit. Exim now panics and dies if it cannot write to a log file, just
6063 as it does if it cannot open a log file.
6065 39. Modified OS/Makefile-Linux so that it now contains
6067 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6069 The two -D definitions ensure that Exim is compiled with large file
6070 support, which makes it possible to handle log files that are bigger than
6073 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6074 instance) a domain was checked against a named list that involved a lookup,
6075 causing $domain_data to be set, then another domain was checked against the
6076 same list, then the first domain was re-checked, the value of $domain_data
6077 after the final check could be wrong. In particular, if the second check
6078 failed, it could be set empty. This bug probably also applied to
6081 41. The strip_trailing_dot option was not being applied to the address given
6082 with the -f command-line option.
6084 42. The code for reading a message's header from the spool was incrementing
6085 $received_count, but never initializing it. This meant that the value was
6086 incorrect (doubled) while delivering a message in the same process in which
6087 it was received. In the most common configuration of Exim, this never
6088 happens - a fresh exec is done - but it can happen when
6089 deliver_drop_privilege is set.
6091 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6092 it now includes up to 150 characters of the unexpected data in the log
6095 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6096 and building data strings. The size of both of these buffers was 10 000
6097 bytes - far larger than anybody would *ever* want, thought I. Needless to
6098 say, somebody hit the limit. I have increased the maximum line length to
6099 20 000 and the maximum data length of concatenated lines to 100 000. I have
6100 also fixed two bugs, because there was no checking on these buffers. Tsk,
6101 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6102 buffer is too small.
6104 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6105 lsearch lookups. Now it does.
6107 46. When parsing a route_list item in a manualroute router, a fixed-length
6108 buffer was used for the list of hosts. I made this 1024 bytes long,
6109 thinking that nobody would ever have a list of hosts that long. Wrong.
6110 Somebody had a whole pile of complicated expansion conditions, and the
6111 string was silently truncated, leading to an expansion error. It turns out
6112 that it is easier to change to an unlimited length (owing to other changes
6113 that have happened since this code was originally written) than to build
6114 structure for giving a limitation error. The length of the item that
6115 expands into the list of hosts is now unlimited.
6117 47. The lsearch lookup could not handle data where the length of text line was
6118 more than 4095 characters. Such lines were truncated, leading to shortened
6119 data being returned. It should now handle lines of any length.
6121 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6122 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6125 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6126 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6127 no longer happens in comment lines. A list of replacements is now placed
6128 at the head of all of the source files, except those whose only change is
6129 to replace PERL_COMMAND in the very first #! line.
6131 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6132 messages on the queue, with a bottom-up merge sort, using code contributed
6133 by Michael Haardt. This should make operations like -bp somewhat faster on
6134 large queues. It won't affect queue runners, except when queue_run_in_order
6137 51. Installed eximstats 1.31 in the distribution.
6139 52. Added support for SRV lookups to the dnslookup router.
6141 53. If an ACL referred to $message_body or $message_body_end, the value was not
6142 reset for any messages that followed in the same SMTP session.
6144 54. The store-handling optimization for building very long strings was not
6145 differentiating between the different store pools. I don't think this
6146 actually made any difference in practice, but I've tidied it.
6148 55. While running the routers to verify a sender address, $sender_address
6149 was still set to the sender address. This is wrong, because when routing to
6150 send a bounce to the sender, it would be empty. Therefore, I have changed
6151 it so that, while verifying a sender address, $sender_address is set to <>.
6152 (There is no change to what happens when verifying a recipient address.)
6154 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6155 target A or AAAA records (if not already returned) without resetting the
6156 qualify_single or search_parents options of the DNS resolver. These are
6157 inappropriate in this case because the targets of MX and SRV records must
6158 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6159 target that, when qualified, matched something in the local domain. These
6160 two options are now turned off when doing these lookups.
6162 57. It seems that at least some releases of Reiserfs (which does not have the
6163 concept of a fixed number of inodes) returns zero and not -1 for the
6164 number of available inodes. This interacted badly with check_spool_inodes,
6165 which assumed that -1 was the "no such thing" setting. What I have done is
6166 to check that the total number of inodes is greater than zero before doing
6167 the test of how many are available.
6169 58. When a "warn" ACL statement has a log_message modifier, the message is
6170 remembered, and not repeated. This is to avoid a lot of repetition when a
6171 message has many recipients that cause the same warning to be written.
6172 However, Exim was preserving the list of already written lines for an
6173 entire SMTP session, which doesn't seem right. The memory is now reset if a
6174 new message is started.
6176 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6177 debugging output unless log_rewrite was also set.
6179 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6180 of (int)(handle) when we know that handle contains (void *)(-1).
6182 61. The Exim daemon panic-logs an error return when it closes the incoming
6183 connection. However "connection reset by peer" seems to be common, and
6184 isn't really an error worthy of noting specially, so that particular error
6187 62. When Exim is trying to find all the local interfaces, it used to panic and
6188 die if the ioctl to get the interface flags failed. However, it seems that
6189 on at least one OS (Solaris 9) it is possible to have an interface that is
6190 included in the list of interfaces, but for which you get a failure error
6191 for this call. This happens when the interface is not "plumbed" into a
6192 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6193 failure of the "get flags" call assumes that the interface is down.
6195 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6196 makes life easier for people who are doing arithmetic on fields extracted
6197 from dates, where you often get leading zeros that should not be
6198 interpreted as octal.
6200 64. Added qualify_domain to the redirect router, to override the global
6203 65. If a pathologically long header line contained very many addresses (the
6204 report of this problem mentioned 10 000) and each of them was rewritten,
6205 Exim could use up a very large amount of memory. (It kept on making new
6206 copies of the header line as it rewrote, and never released the old ones.)
6207 At the expense of a bit more processing, the header rewriting function has
6208 been changed so that it no longer eats memory in this way.
6210 66. The generation of the Received: header has been moved from the time that a
6211 message starts to be received, to the time that it finishes. The timestamp
6212 in the Received: header should now be very close to that of the <= log
6213 line. There are two side-effects of this change:
6215 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6216 logged header lines no longer include the local Received: line, because
6217 it has not yet been created. The same applies to a copy of the message
6218 that is returned to a non-SMTP sender when a message is rejected.
6220 (b) When a filter file is tested using -bf, no additional Received: header
6221 is added to the test message. After some thought, I decided that this
6224 This change does not affect the value of $received_for. It is still set
6225 after address rewriting, but before local_scan() is called.
6227 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6229 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6230 gave an unhelpful panic error message, and a defer error. I have managed to
6231 change this behaviour so that it now rejects any supplied certificate,
6232 which seems right, as the list of acceptable certificates is empty.
6234 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6235 gave an unhelpful defer error. I have not managed to make this reject any
6236 supplied certificates, but the error message it gives is "no certificate
6237 supplied", which is not helpful.
6239 70. exigrep's output now also includes lines that are not associated with any
6240 message, but which match the given pattern. Implemented by a patch from
6241 Martin Sluka, which also tidied up the Perl a bit.
6243 71. Recipient callout verification, like sender verification, was using <> in
6244 the MAIL FROM command. This isn't really the right thing, since the actual
6245 sender may affect whether the remote host accepts the recipient or not. I
6246 have changed it to use the actual sender in the callout; this means that
6247 the cache record is now keyed on a recipient/sender pair, not just the
6248 recipient address. There doesn't seem to be a real danger of callout loops,
6249 since a callout by the remote host to check the sender would use <>.
6250 [SEE ABOVE: changed after hitting problems.]
6252 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6253 temporary errors. However, in the case of such a code being given after
6254 the end of a data transmission (i.e. after ".") Exim was failing to write
6255 a retry record for the message. (Yes, there was some broken host that was
6256 actually sending 8xx at this point.)
6258 73. An unknown lookup type in a host list could cause Exim to panic-die when
6259 the list was checked. (An example that provoked this was putting <; in the
6260 middle of a list instead of at the start.) If this happened during a DATA
6261 ACL check, a -D file could be left lying around. This kind of configuration
6262 error no longer causes Exim to die; instead it causes a defer error. The
6263 incident is still logged to the main and panic logs.
6265 74. Buglet left over from Exim 3 conversion. The message "too many messages
6266 in one connection" was written to the rejectlog but not the mainlog, except
6267 when address rewriting (yes!) was being logged.
6269 75. Added write_rejectlog option.
6271 76. When a system filter was run not as root (that is, when system_filter_user
6272 was set), the values of the $n variables were not being returned to the
6273 main process; thus, they were not subsequently available in the $sn
6276 77. Added +return_path_on_delivery log selector.
6278 78. A connection timeout was being treated differently from recipients deferred
6279 when testing hosts_max_try with a message that was older than the host's
6280 retry timeout. (The host should not be counted, thus allowing all hosts to
6281 be tried at least once before bouncing.) This may have been the cause of an
6282 occasionally reported bug whereby a message would remain on the queue
6283 longer than the retry timeout, but would be bounced if a delivery was
6284 forced. I say "may" because I never totally pinned down the problem;
6285 setting up timeout/retry tests is difficult. See also the next item.
6287 79. The ultimate address timeout was not being applied to errors that involved
6288 a combination of host plus message (for example, a timeout on a MAIL
6289 command). When an address resolved to a number of possible hosts, and they
6290 were not all tried for each delivery (e.g. because of hosts_max_try), a
6291 message could remain on the queue longer than the retry timeout.
6293 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6296 81. Fixed an obscure SMTP outgoing bug which required at least the following
6297 conditions: (a) there was another message waiting for the same server;
6298 (b) the server returned 5xx to all RCPT commands in the first message so
6299 that the message was not completed; (c) the server dropped the connection
6300 or gave a negative response to the RSET that Exim sends to abort the
6301 transaction. The observed case was a dropped connection after DATA that had
6302 been sent in pipelining mode. That is, the server had advertised PIPELINING
6303 but was not implementing it correctly. The effect of the bug was incorrect
6304 behaviour, such as trying another host, and this could lead to a crash.
6310 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6311 and daemon.c were passed as pointers to ints; they should have been
6312 pointers to socklen_t variables (which are typically unsigned ints).
6314 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6317 3. Fixed a really odd bug that affected only the testing scheme; patching a
6318 certain fixed string in the binary changed the value of another string that
6319 happened to be identical to the end of the original first string.
6321 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6322 name", it returns that address as the IP address. On some operating
6323 systems (e.g. Solaris), it also passes back the IP address string as the
6324 "host name". However, on others (e.g. Linux), it passes back an empty
6325 string. Exim wasn't checking for this, and was changing the host name to an
6326 empty string, assuming it had been canonicalized.
6328 5. Although rare, it is permitted to have more than one PTR record for a given
6329 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6330 all the names associated with an address, because they do in Solaris.
6331 However, it seems that they do not in Linux for data that comes from the
6332 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6333 I found this out when I moved to a new Linux workstation and tried to run
6334 the Exim test suite.
6336 To get round this problem I have changed the code so that it now does its
6337 own call to the DNS to look up PTR records when searching for a host name.
6338 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6339 addresses that are only in /etc/hosts are still found.
6341 This behaviour is, however, controlled by an option called host_lookup_
6342 order, which defaults to "bydns:byaddr". If people want to use the other
6343 order, or indeed, just use one or the other means of lookup, they can
6344 specify it in this variable.
6346 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6347 some operating systems, this comes back from gethostbyaddr() as an empty
6348 string, and this is what Exim used to test for. However, it seems that in
6349 other systems, "." is yielded. Exim now tests for this case too.
6351 7. The values of check_spool_space and check_log_space are now held internally
6352 as a number of kilobytes instead of an absolute number of bytes. If a
6353 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6354 kilobyte. This means that much larger values can be stored.
6356 8. Exim monitor: an attempt to get the action menu when not actually pointing
6357 at a message produces an empty menu entitled "No message selected". This
6358 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6359 no entries in it ("Shell widget menu has zero width and/or height"). So I
6360 have added a single, blank menu entry in this case.
6362 9. Added ${quote_local_part.
6364 10. MIME decoding is now applied to the contents of Subject: header lines when
6367 11. Now that a reference to $sender_host_address automatically causes a reverse
6368 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6369 host lookup before query-style lookups in lists that might use this
6370 variable. This has therefore been abolished, and the "net-" prefix is no
6371 longer necessary for query-style lookups.
6373 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6374 have been a typo for LFLAGS, so it has been changed.
6376 13. The install script calls Exim with "-C /dev/null" in order to find the
6377 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6378 to be output. However, since Exim outputs its version number before the
6379 error, it didn't break the script. It just looked ugly. I fixed this by
6380 always allowing "-C /dev/null" if the caller is root.
6382 14. Ignore overlarge ACL variable number when reading spool file - insurance
6383 against a later release with more variables having written the file.
6385 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6386 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6387 this, as well as the form without the "IPv6" on the front.
6389 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6390 OS/Makefile-Darwin file.
6392 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6393 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6395 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6396 variable. However, some OS use ints for inodes. I've added cast to long int
6397 to get rid of the compiler warning.
6399 19. I had forgotten to lock out "/../" in configuration file names when
6400 ALT_CONFIG_PREFIX was set.
6402 20. Routers used for verification do not need to specify transports. However,
6403 if such a router generated a host list, and callout was configured, Exim
6404 crashed, because it could not find a port number from the (non-existent)
6405 transport. It now assumes port 25 in this circumstance.
6407 21. Added the -t option to exigrep.
6409 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6410 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6411 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6412 them is set, LOOKUP_LSEARCH is forced.
6414 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6416 24. Added sender and host information to the "rejected by local_scan()" log
6417 line; previously there was no indication of these.
6419 25. Added .include_if_exists.
6421 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6422 Linux. It turns out that not all file systems support this. Apparently some
6423 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6424 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6425 not supported on the file descriptor, is now ignored when Exim is trying to
6426 sync a directory. This applies only to Linux.
6428 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6430 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6431 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6434 29. Added support for maildirsize files from supplied patch (modified a bit).
6436 30. The use of :fail: followed by an empty string could lead Exim to respond to
6437 sender verification failures with (e.g.):
6439 550 Verification failed for <xxx>
6440 550 Sender verify failed
6442 where the first response line was missing the '-' that indicates it is not
6443 the final line of the response.
6445 31. The loop for finding the name of the user that called Exim had a hardwired
6446 limit of 10; it now uses the value of finduser_retries, which is used for
6447 all other user lookups.
6449 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6452 33. Exim was neglecting to zero errno before one call of strtol() when
6453 expanding a string and expecting an integer value. On some systems this
6454 resulted in spurious "integer overflow" errors. Also, it was casting the
6455 result into an int without checking.
6457 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6458 did not work. The code looks as if it has *never* worked, though it appears
6459 to have been documented since at least release 1.62. I have made it work.
6461 35. The "timeout_DNS" error in retry rules, also documented since at least
6462 1.62, also never worked. As it isn't clear exactly what this means, and
6463 clearly it isn't a major issue, I have abolished the feature by treating it
6464 as "timeout", and writing a warning to the main and panic logs.
6466 36. The display of retry rules for -brt wasn't always showing the error code
6469 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6470 timeout_connect_A, timeout_connect_MX.
6472 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6473 to the empty sender.
6475 39. The daemon was not analysing the content of -oX till after it had closed
6476 stderr and disconnected from the controlling terminal. This meant that any
6477 syntax errors were only noted on the panic log, and the return code from
6478 the command was 0. By re-arranging the code a little, I've made the
6479 decoding happen first, so such errors now appear on stderr, and the return
6480 code is 1. However, the actual setting up of the sockets still happens in
6481 the disconnected process, so errors there are still only recorded on the
6484 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6485 connections (as happens on some IP stacks) was logged at start up time as
6486 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6487 it from "IPv6 and IPv4", which means that two separate sockets are being
6490 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6491 says whether AF_INET or AF_INET6 was passed as an argument.
6493 42. Exiwhat output was messed up when time zones were included in log
6496 43. Exiwhat now gives more information about the daemon's listening ports,
6497 and whether -tls-on-connect was used.
6499 44. The "port" option of the smtp transport is now expanded.
6501 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6502 silently ignored. Now an error message is written to the main and panic
6505 46. There's a new ACL modifier called "logwrite" which writes to a log file
6506 as soon as it is encountered.
6508 47. Added $local_user_uid and $local_user_gid at routing time.
6510 48. Exim crashed when trying to verify a sender address that was being
6513 49. Exim was recognizing only a space character after ".include". It now also
6514 recognizes a tab character.
6516 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6517 extracting the relevant information from the specification. The man page no
6518 longer contains scrambled data for the -d option, and I've added a section
6519 at the front about calling Exim under different names.
6521 51. Added "extra_headers" argument to the "mail" command in filter files.
6523 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6526 53. Installed eximstats 1.29.
6528 54. Added transport_filter_timeout as a generic transport option.
6530 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6531 Cc: header lines. This was required by RFC 822, but it not required by RFC
6534 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6535 incoming messages that did not have them. Now it does so only if the
6536 message originates locally, that is, if there is no associated remote host
6537 address. When Resent- header lines are present, this applies to the Resent-
6538 lines rather than the non-Resent- lines.
6540 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6541 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6543 58. Messages for configuration errors now include the name of the main
6544 configuration file - useful now that there may be more than one file in a
6545 list (.included file names were always shown).
6547 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6548 for those rare installations that do not start the daemon as root or run it
6549 setuid root. I've cut out the call to initgroups() if the daemon is not
6552 60. The Exim user and group can now be bound into the binary as text strings
6553 that are looked up at the start of Exim's processing.
6555 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6557 62. Added $mailstore_basename variable.
6559 63. Installed patch to sieve.c from Michael Haardt.
6561 64. When Exim failed to open the panic log after failing to open the main log,
6562 the original message it was trying to log was written to stderr and debug
6563 output, but if they were not available (the usual case in production), it
6564 was lost. Now it is written to syslog before the two lines that record the
6565 failures to open the logs.
6567 65. Users' Exim filters run in subprocesses under the user's uid. It is
6568 possible for a "deliver" command or an alias in a "personal" command to
6569 provoke an address rewrite. If logging of address rewriting is configured,
6570 this fails because the process is not running as root or exim. There may be
6571 a better way of dealing with this, but for the moment (because 4.30 needs
6572 to be released), I have disabled address rewrite logging when running a
6573 filter in a non-root, non-exim process.
6579 1. The buildconfig auxiliary program wasn't quoting the value set for
6580 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6581 not defined. This bug was masked in 4.22 by the effect that was fixed in
6584 2. Some messages that were rejected after a message id was allocated were
6585 shown as "incomplete" by exigrep. It no longer does this for messages that
6586 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6588 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6589 have allow_domain_literals set, the ID did not get logged in the <= line.
6590 Domain literals are now always recognized in Message-ID: header lines.
6592 4. The first argument for a ${extract expansion item is the key name or field
6593 number. Leading and trailing spaces in this item were not being ignored,
6594 causing some misleading effects.
6596 5. When deliver_drop_privilege was set, single queue runner processes started
6597 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6598 same command in the process it spins off) were not dropping privilege.
6600 6. When the daemon running as "exim" started a queue runner, it always
6601 re-executed Exim in the spun-off process. This is a waste of effort when
6602 deliver_drop_privilege is set. The new process now just calls the
6603 queue-runner function directly.
6609 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6612 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6613 ignored. Though the use of -oP was forcing the writing of a pid file, it
6614 was always written to the default place.
6616 3. If the message "no IP address found for host xxxx" is generated during
6617 incoming verification, it is now followed by identification of the incoming
6618 connection (so you can more easily find what provoked it).
6620 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6622 5. Added some features to "harden" Exim a bit more against certain attacks:
6624 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6625 be put in Local/Makefile. This is like the never_users runtime option,
6626 but it cannot be overridden. The default setting is "root".
6628 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6629 prefix string with which any file named in a -C command line option
6632 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6633 is retained for -C and -D only if the caller of Exim is root. Without
6634 it, the exim user may also use -C and -D and retain privilege.
6636 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6637 command line option is disabled.
6639 6. Macro names set by the -D option must start with an upper case letter, just
6640 like macro names defined in the configuration file.
6642 7. Added "dereference=" facility to LDAP.
6644 8. Two instances of the typo "uknown" in the source files are fixed.
6646 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6647 the Configure-Makefile script screwed up while processing it.
6649 10. Incorporated PCRE 4.4.
6651 11. The SMTP synchronization check was not operating right at the start of an
6652 SMTP session. For example, it could not catch a HELO sent before the client
6653 waited for the greeting. There is now a check for outstanding input at the
6654 point when the greeting is written. Because of the duplex, asynchronous
6655 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6656 way, but not yet received, when the check is performed.
6658 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6659 on TCP/IP sockets, because this apparently causes some broken clients to
6662 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6663 unchanged) from the Cygwin maintainer.
6665 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6666 was supported instead of testing for mbx. Effectively a typo.
6668 15. The spa authenticator server code was not checking that the input it
6669 received was valid base64.
6671 16. The debug output line for the "set" modifier in ACLs was not showing the
6672 name of the variable that was being set.
6674 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6676 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6677 lookup on its own. Something else was needed to trigger the lookup. For
6678 example, a match in host_lookup or the need for a host name in a host list.
6679 Now, if $sender_host_name is referenced and the host name has not yet been
6680 looked up, a lookup is performed. If the lookup fails, the variable remains
6681 empty, and $host_lookup_failed is set to "1".
6683 19. Added "eqi" as a case-independent comparison operator.
6685 20. The saslauthd authentication condition could segfault if neither service
6686 nor realm was specified.
6688 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6689 error message that was logged was misleading, and incoming SMTP
6690 connections were dropped. The message is now more accurate, and temporary
6691 errors are given to SMTP connections.
6693 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6694 (except RSET) with a 421 error, until QUIT is received. However, it was
6695 failing to send a response to QUIT.
6697 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6698 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6699 ACL is now run after the helo_try_verify_hosts code.
6701 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6702 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6703 case-independent, so other case variants are also recognized). Apparently
6704 some systems use these upper case variants.
6706 25. If more than two messages were waiting for the same host, and a transport
6707 filter was specified for the transport, Exim sent two messages over the
6708 same TCP/IP connection, and then failed with "socket operation on non-
6709 socket" when it tried to send the third.
6711 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6713 27. The extern definition of crypt16() in expand.c was not being excluded when
6714 the OS had its own crypt16() function.
6716 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6717 as a preferred synonym for return_size_limit, both as an option and as an
6720 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6722 30. Changed the default configuration ACL to relax the local part checking rule
6723 for addresses that are not in any local domains. For these addresses,
6724 slashes and pipe symbols are allowed within local parts, but the sequence
6725 /../ is explicitly forbidden.
6727 31. SPA server authentication was not clearing the challenge buffer before
6730 32. log_message in a "warn" ACL statement was writing to the reject log as
6731 well as to the main log, which contradicts the documentation and doesn't
6732 seem right (because no rejection is happening). So I have stopped it.
6734 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6735 However, I am unable to do any testing of this.
6737 34. Fixed an infelicity in the appendfile transport. When checking directories
6738 for a mailbox, to see if any needed to be created, it was accidentally
6739 using path names with one or more superfluous leading slashes; tracing
6740 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6742 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6743 contents are added to the log line that is written for every discarded
6744 recipient. (Previously a log_message setting was ignored.)
6746 36. The ${quote: operator now quotes the string if it is empty.
6748 37. The install script runs exim in order to find its version number. If for
6749 some reason other than non-existence or emptiness, which it checks, it
6750 could not run './exim', it was installing it with an empty version number,
6751 i.e. as "exim-". This error state is now caught, and the installation is
6754 38. An argument was missing from the function that creates an error message
6755 when Exim fails to connect to the socket for saslauthd authentication.
6756 This could cause Exim to crash, or give a corrupted message.
6758 39. Added isip, isip4, and isip6 to ${if conditions.
6760 40. The ACL variables $acl_xx are now saved with the message, and can be
6761 accessed later in routers, transports, and filters.
6763 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6764 strings in the file are not string-expanded.
6766 42. If a MAIL command specified a SIZE value that was too large to fit into an
6767 int variable, the check against message_size_limit failed. Such values are
6768 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6769 day this will have to be increased, but I don't think I want to be around
6770 when emails are that large.
6777 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6778 iconv() is not standard in FreeBSD.
6780 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6781 IPv6 enabled. The observed symptom was a segmentation fault on return from
6782 the function os_common_find_running_interfaces() in src/os.c.
6784 3. In the check_special_case() function in daemon.c I had used "errno" as an
6785 argument name, which causes warnings on some systems. This was basically a
6786 typo, since it was named "eno" in the comments!
6788 4. The code that waits for the clock to tick (at a resolution of some fraction
6789 of a second) so as to ensure message-id uniqueness was always waiting for
6790 at least one whole tick, when it could have waited for less. [This is
6791 almost certainly not relevant at current processor speeds, where it is
6792 unlikely to ever wait at all. But we try to future-proof.]
6794 5. The function that sleeps for a time interval that includes fractions of a
6795 second contained a race. It did not block SIGALRM between setting the
6796 timer, and suspending (a couple of lines later). If the interval was short
6797 and the sigsuspend() was delayed until after it had expired, the suspension
6798 never ended. On busy systems this could lead to processes getting stuck for
6801 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6802 process, before it forks any delivery processes. The open lookup caching
6803 mechanism meant that the open file or database connection was passed into
6804 the delivery process. The problem was that delivery processes always tidy
6805 up cached lookup data. This could cause a problem for the next delivery
6806 process started by the queue runner, because the external queue runner
6807 process does not know about the closure. So the next delivery process
6808 still has data in the lookup cache. In the case of a file lookup, there was
6809 no problem because closing a file descriptor in a subprocess doesn't affect
6810 the parent. However, if the lookup was caching a connection to a database,
6811 the connection was closed, and the second delivery process was likely to
6812 see errors such as "PGSQL: query failed: server closed the connection
6813 unexpectedly". The problem has been fixed by closing all cached lookups
6814 in a queue runner before running a delivery process.
6816 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6817 seem to have the "const" qualifier which it has on other OS. I've
6820 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6821 *of the same type* (client or server) with the same public name that an
6822 error should be diagnosed.
6824 9. When Exim looked up a host name for an IP address, but failed to find the
6825 original IP address when looking up the host name (a safety check), it
6826 output the message "<ip address> does not match any IP for NULL", which was
6827 confusing, to say the least. The bug was that the host name should have
6828 appeared instead of "NULL".
6830 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6831 user that is built into the binary, and the -C or -D options is used, root
6832 privilege is dropped before the configuration file is read. In addition,
6833 logging is switched to stderr instead of the normal log files. If the
6834 configuration then re-defines the Exim user, the unprivileged environment
6835 is probably not what is expected, so Exim logs a panic warning message (but
6838 However, if deliver_drop_privilege is set, the unprivileged state may well
6839 be exactly what is intended, so the warning has been cut out in that case,
6840 and Exim is allowed to try to write to its normal log files.
6846 1. smtp_return_error_details was not giving details for temporary sender
6847 or receiver verification errors.
6849 2. Diagnose a configuration error if two authenticators have the same public
6852 3. Exim used not to create the message log file for a message until the first
6853 delivery attempt. This could be confusing when incoming messages were held
6854 for policy or load reasons. The message log file is now created at the time
6855 the message is received, and an initial "Received" line is written to it.
6857 4. The automatically generated man page for command line options had a minor
6858 bug that caused no ill effects; however, a more serious problem was that
6859 the procedure for building the man page automatically didn't always
6860 operate. Consequently, release 4.20 contains an out-of-date version. This
6861 shouldn't happen again.
6863 5. When building Exim with embedded Perl support, the script that builds the
6864 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6865 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6867 6. The freeze_tell option was not being used for messages that were frozen on
6868 arrival, either by an ACL or by local_scan().
6870 7. Added the smtp_incomplete_transaction log selector.
6872 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6873 was accepting AUTH without a new EHLO.
6875 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6876 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6879 10. Logging of TCP/IP connections (when configured) now happens in the main
6880 daemon process instead of the child process, so that the TCP/IP connection
6881 count is more accurate (but it can never be perfect).
6883 11. The use of "drop" in a nested ACL was not being handled correctly in the
6884 outer ACL. Now, if condition failure induced by the nested "drop" causes
6885 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6886 or "require"), the connection is dropped.
6888 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6889 that yield "discard" can now be used with an "accept" or a "discard" verb,
6890 but an error is generated for any others (because I can't see a useful way
6891 to define what should happen).
6893 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6894 are now processed in the same way as lines in the Exim configuration file.
6895 In particular, continuation lines are supported.
6897 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6899 15. Added -ti meaning -t -i.
6901 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6902 domains, and warn by logging if others are found.
6904 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6905 returned from GIFCONF when Exim is trying to find the list of interfaces on
6906 a host. The code in os.c has been modified to copy each ifreq to an aligned
6907 structure in all cases.
6909 Also, in some cases, the returned ifreq's were being copied to a 'struct
6910 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6911 means the last couple of bytes of an IPv6 address could be chopped if the
6912 ifreq contained only a normal sockaddr (14 bytes storage).
6914 18. Named domain lists were not supported in the hosts_treat_as_local option.
6915 An entry such as +xxxx was not recognized, and was treated as a literal
6918 19. Ensure that header lines added by a DATA ACL are included in the reject log
6919 if the ACL subsequently rejects the message.
6921 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6922 MD5 (which is deprecated).
6924 21. When testing a filter file using -bf, Exim was writing a message when it
6925 took the sender from a "From " line in the message, but it was not doing so
6926 when it took $return_path from a Return-Path: header line. It now does.
6928 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6929 with a valid header line field name (a series of printing characters
6930 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6932 23. Changed "disc" in the source to "disk" to conform to the documentation and
6933 the book and for uniformity.
6935 24. Ignore Sendmail's -Ooption=value command line item.
6937 25. When execve() failed while trying to run a command in a pipe transport,
6938 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6939 could be confused with a return value of 69 from the command itself. This
6940 has been changed to 127, the value the shell returns if it is asked to run
6941 a non-existent command. The wording for the related log line suggests a
6942 non-existent command as the problem.
6944 26. If received_header_text expands to an empty string, do not add a Received:
6945 header line to the message. (Well, it adds a token one on the spool, but
6946 marks it "old" so that it doesn't get used or transmitted.)
6948 27. Installed eximstats 1.28 (addition of -nt option).
6950 28. There was no check for failure on the call to getsockname() in the daemon
6951 code. This can fail if there is a shortage of resources on the system, with
6952 ENOMEM, for example. A temporary error is now given on failure.
6954 29. Contrary to the C standard, it seems that in some environments, the
6955 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6956 program. Exim now does this explicitly; it affects the formatting of
6957 timestamps using strftime().
6959 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6960 complaints. I've now initialized all the variables, to avoid this.
6962 32. Header lines added by a system filter were not being "seen" during
6963 transport-time rewrites.
6965 33. The info_callback() function passed to OpenSSL is set up with type void
6966 (*)(SSL *, int, int), as described somewhere. However, when calling the
6967 function (actually a macro) that sets it up, the type void(*)() is
6968 expected. I've put in a cast to prevent warnings from picky compilers.
6970 34. If a DNS black list lookup found a CNAME record, but there were no A
6971 records associated with the domain it pointed at, Exim crashed.
6973 35. If a DNS black list lookup returned more than one A record, Exim ignored
6974 all but the first. It now scans all returned addresses if a particular IP
6975 value is being sought. In this situation, the contents of the
6976 $dnslist_value variable are a list of all the addresses, separated by a
6979 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6980 to accept a host name and all its aliases if the forward lookup for any of
6981 them yielded the IP address of the incoming connection. Now it accepts only
6982 those names whose forward lookup yields the correct IP address. Any other
6983 names are discarded. This closes a loophole whereby a rogue DNS
6984 administrator could create reverse DNS records to break through a
6985 wildcarded host restriction in an ACL.
6987 37. If a user filter or a system filter that ran in a subprocess used any of
6988 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6989 the wrong values were passed to the pipe command ($thisaddress had the
6990 value of $0, $0 had the value of $1, etc). This bug was introduced by
6991 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6993 38. Improved the line breaking for long SMTP error messages from ACLs.
6994 Previously, if there was no break point between 40 and 75 characters, Exim
6995 left the rest of the message alone. Two changes have been made: (a) I've
6996 reduced the minimum length to 35 characters; (b) if it can't find a break
6997 point between 35 and 75 characters, it looks ahead and uses the first one
6998 that it finds. This may give the occasional overlong line, but at least the
6999 remaining text gets split now.
7001 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7002 file descriptors might be low, and that setting 1000 would always raise it.
7003 It turns out that in some environments, the limit is already over 1000 and
7004 that lowering it causes trouble. So now Exim takes care not to decrease it.
7006 40. When delivering a message, the value of $return_path is set to $sender_
7007 address at the start of routing (routers may change the value). By an
7008 oversight, this default was not being set up when an address was tested by
7009 -bt or -bv, which affected the outcome if any router or filter referred to
7012 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7013 log only when "message" or "log_message" are set. However, if one of the
7014 conditions was an address verification, or a call to a nested ACL, the
7015 messages generated by the underlying test were being passed through. This
7016 no longer happens. The underlying message is available in $acl_verify_
7017 message for both "message" and "log_message" expansions, so it can be
7018 passed through if needed.
7020 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7021 new expansion $bh_ to give the encoded byte string without charset
7022 translation. Translation happens only if iconv() is available; HAVE_ICONV
7023 indicates this at build time. HEADERS_CHARSET gives the charset to
7024 translate to; headers_charset can change it in the configuration, and
7025 "headers charset" can change it in an individual filter file.
7027 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7028 that creates RFC 2047 encoded "words" labels them as that charset instead
7029 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7030 expansion operator; (ii) when Exim creates a From: line for a local
7031 message; (iii) when a header line is rewritten to include a "phrase" part.
7033 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7034 buggy, causing it to skip the first lines of messages whose message ID
7035 ended in 'D'. This would not have bitten before Exim release 4.14, because
7036 message IDs were unlikely to end in 'D' before then. The effect was to have
7037 incorrect size information for certain domains.
7039 45. #include "config.h" was missing at the start of the crypt16.c module. This
7040 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7043 46. If there was a timeout during a "random" callout check, Exim treated it as
7044 a failure of the random address, and carried on sending RSET and the real
7045 address. If the delay was just some slowness somewhere, the response to the
7046 original RCPT would be taken as a response to RSET and so on, causing
7047 mayhem of various kinds.
7049 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7050 when I implemented it. It didn't allow for the fact that some option values
7051 may legitimately be negative (e.g. size_addition), and it didn't even do
7052 the right test for positive values.
7054 48. Domain names in DNS records are case-independent. Exim always looks them up
7055 in lower case. Some resolvers return domain names in exactly the case they
7056 appear in the zone file, that is, they may contain uppercase letters. Not
7057 all resolvers do this - some return always lower case. Exim was treating a
7058 change of case by a resolver as a change of domain, similar to a widening
7059 of a domain abbreviation. This triggered its re-routing code and so it was
7060 trying to route what was effectively the same domain again. This normally
7061 caused routing to fail (because the router wouldn't handle the domain
7062 twice). Now Exim checks for this case specially, and just changes the
7063 casing of the domain that it ultimately uses when it transmits the message
7066 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7069 50. If a filter generated a file delivery with a non-absolute name (possible if
7070 no home directory exists for the router), the forbid_file option was not
7073 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7074 the existing equality matching.
7076 52. Exim was using ints instead of ino_t variables in some places where it was
7077 dealing with inode numbers.
7079 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7080 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7081 called TMPDIR, and if it finds it is different, it changes its value.
7083 54. The smtp_printf() function is now made available to local_scan() so
7084 additional output lines can be written before returning. There is also an
7085 smtp_fflush() function to enable the detection of a dropped connection.
7086 The variables smtp_input and smtp_batched_input are exported to
7089 55. Changed the default runtime configuration: the message "Unknown user"
7090 has been removed from the ACL, and instead placed on the localuser router,
7091 using the cannot_route_message feature. This means that any verification
7092 failures that generate their own messages won't get overridden. Similarly,
7093 the "Unrouteable address" message that was in the ACL for unverifiable
7094 relay addresses has also been removed.
7096 56. Added hosts_avoid_esmtp to the smtp transport.
7098 57. The exicyclog script was not checking for the esoteric option
7099 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7100 will work only if exicyclog is run under the appropriate euid.
7102 58. Following a discussion on the list, the rules by which Exim recognises line
7103 endings on incoming messages have been changed. The -dropcr and drop_cr
7104 options are now no-ops, retained only for backwards compatibility. The
7105 following line terminators are recognized: LF CRLF CR. However, special
7106 processing applies to CR:
7108 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7109 nor a local message in the state where . is a terminator.
7111 (ii) If a bare CR is encountered in a header line, an extra space is added
7112 after the line terminator so as not to end the header. The reasoning
7113 behind this is that bare CRs in header lines are most likely either
7114 to be mistakes, or people trying to play silly games.
7116 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7117 was being incorrectly given as 18 bytes larger than it should have been.
7118 This is a VOB (very old bug).
7120 60. This may never have affected anything current, but just in case it has:
7121 When the local host is found other than at the start of a list of hosts,
7122 the local host, those with the same MX, and any that follow, are discarded.
7123 When the list in question was part of a longer list of hosts, the following
7124 hosts (not currently being processed) were also being discarded. This no
7125 longer happens. I'm not sure if this situation could ever has previously
7128 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7131 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7132 header. This is something that is recommended in a new Internet Draft, and
7133 is something that is documented as being done by Sendmail. There are two
7134 possible values. For messages generated by the autoreply transport, Exim
7137 Auto-Submitted: auto-replied
7139 whereas for all other generated messages (e.g. bounces) it adds
7141 Auto-Submitted: auto-generated
7143 63. The "personal" condition in filters now includes a test for the
7144 Auto-Submitted: header. If it contains the string "auto-" the message it
7145 not considered personal.
7147 64. Added rcpt_include_affixes as a generic transport option.
7149 65. Added queue_only_override (default true).
7151 66. Added the syslog_duplication option.
7153 67. If what should have been the first header line of a message consisted of
7154 a space followed by a colon, Exim was mis-interpreting it as a header line.
7155 It isn't of course - it is syntactically invalid and should therefore be
7156 treated as the start of the message body. The misbehaviour could have
7157 caused a number of strange effects, including loss of data in subsequent
7158 header lines, and spool format errors.
7160 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7161 client host had authenticated. This control can now be exercised by an ACL
7162 for more flexibility.
7164 69. By default, callouts do not happen when testing with -bh. There is now a
7165 variant, -bhc, which does actually run the callout code, including
7166 consulting and updating the callout cache.
7168 70. Added support for saslauthd authentication, courtesy of Alexander
7171 71. If statvfs() failed on the spool or log directories while checking their
7172 size for availability, Exim confusingly gave the error "space shortage".
7173 Furthermore, in debugging mode it crashed with a floating point exception.
7174 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7175 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7176 serious problem, Exim now writes to the main and panic logs when this
7177 happens, with details of the failure. It then refuses to accept the
7178 incoming message, giving the message "spool directory problem" or "log
7179 directory problem" with a 421 code for SMTP messages.
7181 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7182 0, 1, and 2 exist, because some OS complain for execs without them (see
7183 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7184 descriptors. However, the code omitted to check that the open succeeded,
7185 causing mysterious errors if for some reason the permissions on /dev/null
7186 got screwed. Now Exim writes a message to the main and panic logs, and
7187 bombs out if it can't open /dev/null.
7189 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7190 interact so that it is all more flexible. It is supposed to remain
7191 backwards compatible. Also added extra_local_interfaces.
7193 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7194 to bomb out with an assertion failure - to the client this appears as a
7195 connection drop. This problem occurs in the part of the code that was taken
7196 from the Samba project. Fortunately, the assertion is in a very simple
7197 function, so I have fixed this by reproducing the function inline in the
7198 one place where it is called, and arranging for authentication to fail
7199 instead of killing the process with assert().
7201 75. The SPA client code was not working when the server requested OEM rather
7202 than Unicode encoding.
7204 76. Added code to make require_files with a specific uid setting more usable in
7205 the case where statting the file as root fails - usually a non-root-mounted
7206 NFS file system. When this happens and the failure is EACCES, Exim now
7207 forks a subprocess and does the per-uid checking as the relevant uid.
7209 77. Added process_log_path.
7211 78. If log_file_path was not explicitly set, a setting of check_log_space or
7212 check_log_inodes was ignored.
7214 79. If a space check for the spool or log partitions fails, the incident is now
7215 logged. Of course, in the latter case the data may get lost...
7217 80. Added the %p formatting code to string_format() so that it can be used to
7218 print addresses in debug_print(). Adjusted all the address printing in the
7219 debugging in store.c to use %p rather than %d.
7221 81. There was a concern that a line of code in smtp_in.c could overflow a
7222 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7223 initially expressed, the concern was not well-founded, because trailing
7224 spaces are removed early. However, if the trailing spaces were followed by
7225 a NULL, they did not get removed, so the overflow was possible. Two fixes
7228 (a) I re-wrote the offending code in a cleaner fashion.
7229 (b) If an incoming SMTP command contains a NULL character, it is rejected
7232 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7233 runs initgroups(), so that if the Exim user is in any additional groups,
7234 they will be used during message reception.
7240 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob