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/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
31 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
34 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
35 Previously this would segfault.
37 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
40 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
42 JH/16 Fix ARC verification to do AS checks in reverse order.
44 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
45 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
46 Previously the "utf8" would be re-prepended for every additional message.
48 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
49 one parent address was copied, and bogus data was used at delivery-logging
50 time. Either a crash (after delivery) or bogus log data could result.
51 Discovery and analysis by Tim Stewart.
53 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
54 The ignore_error flag wasn't being returned from the filter subprocess so
55 was not set for later routers. Investigation and fix by Matthias Kurz.
61 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
62 When redis_servers is set to a list of > 1 element, and the Redis servers
63 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
64 case of MOVED into a DEFER case instead, thus moving the query onto the
65 next server in the list. For a cluster of N elements, all N servers must
66 be defined in redis_servers.
68 GF/02 Catch and remove uninitialized value warning in exiqsumm
69 Check for existence of @ARGV before looking at $ARGV[0]
71 JH/01 Replace the store_release() internal interface with store_newblock(),
72 which internalises the check required to safely use the old one, plus
73 the allocate and data copy operations duplicated in both (!) of the
76 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
77 modifier. This matches the restriction on the commandline.
79 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
80 Previously only the last row was returned.
82 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
83 we assumed that tags in the header were well-formed, and parsed the
84 element content after inspecting only the first char of the tag.
85 Assumptions at that stage could crash the receive process on malformed
88 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
89 While running the DKIM ACL we operate on the Permanent memory pool so that
90 variables created with "set" persist to the DATA ACL. Also (at any time)
91 DNS lookups that fail create cache records using the Permanent pool. But
92 expansions release any allocations made on the current pool - so a dnsdb
93 lookup expansion done in the DKIM ACL releases the memory used for the
94 DNS negative-cache, and bad things result. Solution is to switch to the
95 Main pool for expansions.
96 While we're in that code, add checks on the DNS cache during store_reset,
97 active in the testsuite.
98 Problem spotted, and debugging aided, by Wolfgang Breyha.
100 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
101 When none of the hosts presented to a transport match an already-open
102 connection, close it and proceed with the list. Previously we would
103 queue the message. Spotted by Lena with Yahoo, probably involving
106 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
107 Previously a spurious "250 OK id=" response was appended to the proper
110 JH/08 The "support for" informational output now, which built with Content
111 Scanning support, has a line for the malware scanner interfaces compiled
112 in. Interface can be individually included or not at build time.
114 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
115 by the template makefile "src/EDITME". The "STREAM" support for an older
116 ClamAV interface method is removed.
118 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
119 rows affected is given instead).
121 JH/11 The runtime Berkeley DB library version is now additionally output by
122 "exim -d -bV". Previously only the compile-time version was shown.
124 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
125 SMTP connection. Previously, when one had more receipients than the
126 first, an abortive onward connection was made. Move to full support for
127 multiple onward connections in sequence, handling cutthrough connection
128 for all multi-message initiating connections.
130 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
131 routers. Previously, a multi-recipient message would fail to match the
132 onward-connection opened for the first recipient, and cause its closure.
134 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
135 a timeout on read on a GnuTLS initiating connection, resulting in the
136 initiating connection being dropped. This mattered most when the callout
137 was marked defer_ok. Fix to keep the two timeout-detection methods
140 JH/15 Relax results from ACL control request to enable cutthrough, in
141 unsupported situations, from error to silently (except under debug)
142 ignoring. This covers use with PRDR, frozen messages, queue-only and
145 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
147 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
148 metadata, resulting in a crash in free().
150 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
151 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
152 Broken also in d185889f4, with init system revamp.
154 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
155 Previously we abruptly closed the connection after reading a malware-
156 found indication; now we go on to read the "scan ok" response line,
159 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
160 ACL. Previously, a crash would result.
162 JH/19 Speed up macro lookups during configuration file read, by skipping non-
163 macro text after a replacement (previously it was only once per line) and
164 by skipping builtin macros when searching for an uppercase lead character.
166 JH/20 DANE support moved from Experimental to mainline. The Makefile control
167 for the build is renamed.
169 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
170 was allocated for every new TLS startup, meaning one per message. Fix
171 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
173 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
174 reported the original. Fix to report (as far as possible) the ACL
175 result replacing the original.
177 JH/23 Fix memory leak during multi-message connections using STARTTLS under
178 OpenSSL. Certificate information is loaded for every new TLS startup,
179 and the resources needed to be freed.
181 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
183 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
186 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
187 DATA response info to the (existing) per-recipient response info for
188 the "C=" log element. It can have useful tracking info from the
189 destination system. Patch from Simon Arlott.
191 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
192 length value. Previously this would segfault.
194 HS/02 Support Avast multiline protoocol, this allows passing flags to
195 newer versions of the scanner.
197 JH/28 Ensure that variables possibly set during message acceptance are marked
198 dead before release of memory in the daemon loop. This stops complaints
199 about them when the debug_store option is enabled. Discovered specifically
200 for sender_rate_period, but applies to a whole set of variables.
201 Do the same for the queue-runner and queue-list loops, for variables set
202 from spool message files. Do the same for the SMTP per-message loop, for
203 certain variables indirectly set in ACL operations.
205 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
206 as a multi-recipient message from a mailinglist manager). The coding had
207 an arbitrary cutoff number of characters while checking for more input;
208 enforced by writing a NUL into the buffer. This corrupted long / fast
209 input. The problem was exposed more widely when more pipelineing of SMTP
210 responses was introduced, and one Exim system was feeding another.
211 The symptom is log complaints of SMTP syntax error (NUL chars) on the
212 receiving system, and refused recipients seen by the sending system
213 (propating to people being dropped from mailing lists).
214 Discovered and pinpointed by David Carter.
216 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
217 replaced by the ${authresults } expansion.
219 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
221 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
222 allows proper process termination in container environments.
224 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
225 Previously the "final dot" had a newline after it; ensure it is CR,LF.
227 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
228 and "err_perm", deprecated since 4.83 when the RFC-defined words
229 "temperror" and "permerror" were introduced.
231 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
232 transport-filters or DKIM-signing. The restriction was lost in the
233 consolidation of verify-callout and delivery SMTP handling.
234 Extend the restriction to also cover ARC-signing.
236 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
237 in defer=pass mode supply a 450 to the initiator. Previously the message
240 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
241 tls_require_ciphers is used as before.
243 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
244 "pass_unscanned". Only tmpfails from the scanner are written to
245 the paniclog, as they may require admin intervention (permission
246 denied, license issues). Other scanner errors (like decompression
247 bombs) do not cause a paniclog entry.
249 JH/36 Fix reinitialisation of DKIM logging variable between messages.
250 Previously it was possible to log spurious information in receive log
253 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
254 triggered odd behaviour from Outlook Express clients.
256 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
259 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
260 since the IETF WG has not yet settled on that versus the original
261 "bare" representation.
263 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
264 Previously the millisecond value corrupted the output.
265 Fix also for syslog_pid=no and log_selector +pid, for which the pid
266 corrupted the output.
272 JH/01 Rework error string handling in TLS interface so that the caller in
273 more cases is responsible for logging. This permits library-sourced
274 string to be attached to addresses during delivery, and collapses
275 pairs of long lines into single ones.
277 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
278 during configuration. Wildcards are allowed and expanded.
280 JH/02 Rework error string handling in DKIM to pass more info back to callers.
281 This permits better logging.
283 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
284 do not close it down and have the child transport start it up again on
285 the passed-on TCP connection. Instead, proxy the child (and any
286 subsequent ones) for TLS via a unix-domain socket channel. Logging is
287 affected: the continued delivery log lines do not have any DNSSEC, TLS
288 Certificate or OCSP information. TLS cipher information is still logged.
290 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
291 identical IP addresses on different listening ports. Will also affect
294 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
295 add noisy ifdef guards to special-case this sillyness.
296 Patch from Bernd Kuhls.
298 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
299 than 255 are no longer allowed.
301 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
302 Disable the session-cache too, which might reduce our load. Since we
303 currrectly use a new context for every connection, both as server and
304 client, there is no benefit for these.
305 GnuTLS appears to not support tickets server-side by default (we don't
306 call gnutls_session_ticket_enable_server()) but client side is enabled
307 by default on recent versions (3.1.3 +) unless the PFS priority string
310 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
311 <https://reproducible-builds.org/specs/source-date-epoch/>.
313 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
314 the check for any unsuccessful recipients did not notice the limit, and
315 erroneously found still-pending ones.
317 JH/08 Pipeline CHUNKING command and data together, on kernels that support
318 MSG_MORE. Only in-clear (not on TLS connections).
320 JH/09 Avoid using a temporary file during transport using dkim. Unless a
321 transport-filter is involved we can buffer the headers in memory for
322 creating the signature, and read the spool data file once for the
323 signature and again for transmission.
325 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
326 4.77 as the kernel support then wasn't solid, having issues in 64bit
327 mode. Now, it's been long enough. Add support for FreeBSD also.
329 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
330 case where the routing stage had gathered several addresses to send to
331 a host before calling the transport for the first, we previously failed
332 to close down TLS in the old transport process before passing the TCP
333 connection to the new process. The new one sent a STARTTLS command
334 which naturally failed, giving a failed delivery and bloating the retry
335 database. Investigation and fix prototype from Wolfgang Breyha.
337 JH/12 Fix check on SMTP command input synchronisation. Previously there were
338 false-negatives in the check that the sender had not preempted a response
339 or prompt from Exim (running as a server), due to that code's lack of
340 awareness of the SMTP input buffering.
342 PP/04 Add commandline_checks_require_admin option.
343 Exim drops privileges sanely, various checks such as -be aren't a
344 security problem, as long as you trust local users with access to their
345 own account. When invoked by services which pass untrusted data to
346 Exim, this might be an issue. Set this option in main configuration
347 AND make fixes to the calling application, such as using `--` to stop
350 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
351 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
352 responses to those, into a single TLS record each way (this usually means
353 a single packet). As a side issue, smtp_enforce_sync now works on TLS
356 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
357 affects you only if you're dancing at the edge of the param size limits.
358 If you are, and this message makes sense to you, then: raise the
359 configured limit or use OpenSSL 1.1. Nothing we can do for older
362 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
363 cmdline element to get the documented default one. Previously it was
366 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
367 are unsuable for non-message-specific reasons - eg. connection timeout,
368 banner-time rejection.
370 JH/16 Fix logging of delivery remote port, when specified by router, under
373 PP/06 Repair manualroute's ability to take options in any order, even if one
374 is the name of a transport.
377 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
379 JH/17 Change the list-building routines interface to use the expanding-string
380 triplet model, for better allocation and copying behaviour.
382 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
383 Previously it was constructed the first time a possibly-matching string
384 was met in the configuration file input during startup; now it is done
387 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
388 compatible one, to avoid the (poorly documented) possibility of a config
389 file in the working directory redirecting the DB files, possibly correpting
390 some existing file. CVE-2017-10140 assigned for BDB.
392 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
393 cache-hot. Previously, although the result was properly cached, the
394 initial verify call returned a defer.
396 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
397 the main verify for receipient in uncached-mode.
399 JH/22 Retire historical build files to an "unsupported" subdir. These are
400 defined as "ones for which we have no current evidence of testing".
402 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
403 if present. Previously it was ignored.
405 JH/24 Start using specified-initialisers in C structure init coding. This is
406 a C99 feature (it's 2017, so now considered safe).
408 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
409 if was a fixed-sized field and bitmask ops via macros; it is now more
412 PP/07 GitHub PR 56: Apply MariaDB build fix.
413 Patch provided by Jaroslav Škarvada.
415 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
416 during Coverity cleanups [4.87 JH/47]
417 Diagnosis and fix provided by Michael Fischer v. Mollard.
419 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
420 the right size to place the terminating semicolon on its own folded
421 line, the header hash was calculated to an incorrect value thanks to
422 the (relaxed) space the fold became.
424 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
425 and confused the parent.
427 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
428 which could crash as a result. This could lead to undeliverable messages.
430 JH/28 Logging: "next input sent too soon" now shows where input was truncated
433 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
434 matters on fast-turnover and PID-randomising systems, which were getting
435 out-of-order delivery.
437 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
438 a possibly-overlapping copy. The symptom was that "Remote host closed
439 connection in response to HELO" was logged instead of the actual 4xx
442 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
443 Previously only that bufferd was discarded, resulting in SYMTP command
446 JH/32 DKIM: when a message has multiple signatures matching an identity given
447 in dkim_verify_signers, run the dkim acl once for each. Previously only
448 one run was done. Bug 2189.
450 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
451 "panic the current process" to "deliberately defer". The panic log is
452 still written with the problem list name; the mail and reject logs now
453 get a temp-reject line for the message that was being handled, saying
454 something like "domains check lookup or other defer". The SMTP 451
455 message is still "Temporary local problem".
457 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
458 A crafted sequence of BDAT commands could result in in-use memory beeing
459 freed. CVE-2017-16943.
461 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
462 from SMTP input. Previously it was always done; now only done for DATA
463 and not BDAT commands. CVE-2017-16944.
465 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
466 to the message (such as an overlong header line). Previously this was
467 not done and we did not exit BDAT mode. Followon from the previous item
468 though a different problem.
474 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
475 than -2003 did; needs libidn2 in addition to libidn.
477 JH/02 The path option on a pipe transport is now expanded before use.
479 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
480 Patch provided by "Björn", documentation fix added too.
482 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
483 missing a wire-to-host endian conversion.
485 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
486 close after a BDAT command line could be taken as a following command,
487 giving a synch failure. Fix by only checking for synch immediately
488 before acknowledging the chunk.
490 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
491 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
492 macro. Patches provided by Josh Soref.
494 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
495 Previously we did not; the RFC seems ambiguous and VRFY is not listed
496 by IANA as a service extension. However, John Klensin suggests that we
499 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
500 the dkim code may be unix-mode line endings rather than smtp wire-format
501 CRLF, so prepend a CR to any bare LF.
503 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
504 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
506 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
507 the first were themselves being wrongly included in the feed into dkim
508 processing; with most chunk sizes in use this resulted in an incorrect
509 body hash calculated value.
511 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
512 DKIM signature block, for verification. Although advised against by
513 standards it is specifically not ruled illegal.
515 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
517 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
518 missing a body hash (the bh= tag).
520 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
521 It seems that HAProxy sends the Proxy Protocol information in clear and
522 only then does a TLS startup, so do the same.
524 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
525 TCP connections (such as for Spamd) unless the daemon successfully set
526 Fast Open mode on its listening sockets. This fixes breakage seen on
527 too-old kernels or those not configured for Fast Open, at the cost of
528 requiring both directions being enabled for TFO, and TFO never being used
529 by non-daemon-related Exim processes.
531 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
532 endings, at least on the first header line. Try to canonify any that get
533 past that check, despite the cost.
535 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
536 now limited to an arbitrary five deep, while parsing addresses with the
537 strip_excess_angle_brackets option enabled.
539 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
540 instead leave the unprompted TLS handshake in socket buffer for the
541 TLS library to consume.
543 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
545 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
547 JH/16 Drop variables when they go out of scope. Memory management drops a whole
548 region in one operation, for speed, and this leaves assigned pointers
549 dangling. Add checks run only under the testsuite which checks all
550 variables at a store-reset and panics on a dangling pointer; add code
551 explicitly nulling out all the variables discovered. Fixes one known
552 bug: a transport crash, where a dangling pointer for $sending_ip_address
553 originally assigned in a verify callout, is re-used.
555 PP/06 Drop '.' from @INC in various Perl scripts.
557 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
559 PP/08 Reduce a number of compilation warnings under clang; building with
560 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
561 should be warning-free.
563 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
565 HS/01 Fix portability problems introduced by PP/08 for platforms where
566 realloc(NULL) is not equivalent to malloc() [SunOS et al].
568 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
569 chunk. This allows us to accept broken chunked messages. We need a more
570 general solution here.
572 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
573 already-broken messages in the queue.
575 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
577 JH/19 Fix reference counting bug in routing-generated-address tracking.
583 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
584 supports it and a size is available (ie. the sending peer gave us one).
586 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
587 years of being deprecated). The replacements are the ACLs
588 acl_smtp_mime and acl_not_smtp_mime.
590 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
591 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
592 or even an in-clear connection were permitted. Now, if the host lookup
593 was dnssec and dane was requested then the host is only used if the
594 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
595 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
596 if one fails this test.
597 This means that a poorly-configured remote DNS will make it incommunicado;
598 but it protects against a DNS-interception attack on it.
600 JH/04 Bug 1810: make continued-use of an open smtp transport connection
601 non-noisy when a race steals the message being considered.
603 JH/05 If main configuration option tls_certificate is unset, generate a
604 self-signed certificate for inbound TLS connections.
606 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
607 in rewrites and routers.
609 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
610 and logged a warning sing 4.83; now they are a configuration file error.
612 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
613 (lacking @domain). Apply the same qualification processing as RCPT.
615 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
617 JH/10 Support ${sha256:} applied to a string (as well as the previous
620 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
621 a cutthrough deliver is pending, as we always want to make a connection.
622 This also avoids re-routing the message when later placing the cutthrough
623 connection after a verify cache hit.
624 Do not update it with the verify result either.
626 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
627 when routing results in more than one destination address.
629 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
630 signing (which inhibits the cutthrough capability). Previously only
631 the presence of an option was tested; now an expansion evaluating as
632 empty is permissible (obviously it should depend only on data available
633 when the cutthrough connection is made).
635 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
636 the relevant preceding SMTP command did not note the pipelining mode.
638 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
639 Previously they were not counted.
641 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
642 as one having no matching records. Previously we deferred the message
643 that needed the lookup.
645 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
646 distinguished as "(=".
648 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
649 for missing MX records. Previously it only worked for missing A records.
651 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
653 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
654 after the data-go-ahead and data-ack. Patch from Jason Betts.
656 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
657 even for a "none" policy. Patch from Tony Meyer.
659 JH/22 Fix continued use of a connection for further deliveries. If a port was
660 specified by a router, it must also match for the delivery to be
663 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
664 When one of the recipients of a message was unsuitable for the connection
665 (has no matching addresses), we lost track of needing to mark it
666 deferred. As a result mail would be lost.
668 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
670 JH/25 Decoding ACL controls is now done using a binary search; the source code
671 takes up less space and should be simpler to maintain. Merge the ACL
672 condition decode tables also, with similar effect.
674 JH/26 Fix problem with one_time used on a redirect router which returned the
675 parent address unchanged. A retry would see the parent address marked as
676 delivered, so not attempt the (identical) child. As a result mail would
679 JH/27 Fix a possible security hole, wherein a process operating with the Exim
680 UID can gain a root shell. Credit to http://www.halfdog.net/ for
681 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
684 JH/28 Enable {spool,log} filesystem space and inode checks as default.
685 Main config options check_{log,spool}_{inodes,space} are now
686 100 inodes, 10MB unless set otherwise in the configuration.
688 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
689 Previously it only applied to the main-section connection policy
692 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
694 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
695 by me. Added RFC7919 DH primes as an alternative.
697 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
698 are not in the system include path.
700 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
701 GnuTLS, when a session startup failed (eg because the client disconnected)
702 Exim did stdio operations after fclose. This was exposed by a recent
703 change which nulled out the file handle after the fclose.
705 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
706 signed directly by the cert-signing cert, rather than an intermediate
707 OCSP-signing cert. This is the model used by LetsEncrypt.
709 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
711 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
712 an incoming connection.
714 HS/02 Bug 1802: Do not half-close the connection after sending a request
717 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
718 fallback to "prime256v1".
720 JH/34 SECURITY: Use proper copy of DATA command in error message.
721 Could leak key material. Remotely exploitable. CVE-2016-9963.
727 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
728 and 3.4.4 - once the server is enabled to respond to an OCSP request
729 it does even when not requested, resulting in a stapling non-aware
730 client dropping the TLS connection.
732 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
733 support variable-length bit vectors. No functional change.
735 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
736 The I= interface field on outgoing lines is now after the H= remote
737 host field, same as incoming lines. There is a separate
738 outgoing_interface log selector which allows you to disable the
741 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
742 If not running log_selector +smtp_connection the mainlog would be held
743 open indefinitely after a "too many connections" event, including to a
744 deleted file after a log rotate. Leave the per net connection logging
745 leaving it open for efficiency as that will be quickly detected by the
746 check on the next write.
748 HS/01 Bug 1671: Fix post transport crash.
749 Processing the wait-<transport> messages could crash the delivery
750 process if the message IDs didn't exist for some reason. When
751 using 'split_spool_directory=yes' the construction of the spool
752 file name failed already, exposing the same netto behaviour.
754 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
755 mime_regex ACL conditions.
757 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
758 to DSN fail messages (bounces): remote IP, remote greeting, remote response
759 to HELO, local diagnostic string.
761 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
762 log line to debug. Even when configured with a tls authenticator many
763 client connections are expected to not authenticate in this way, so
764 an authenticate fail is not an error.
766 HS/02 Add the Exim version string to the process info. This way exiwhat
767 gives some more detail about the running daemon.
769 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
770 matter for fast-change records such as DNSBLs.
772 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
773 retry record, even if constant. There may be multiple transports with
774 different interface settings and the retry behaviour needs to be kept
777 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
779 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
781 JH/10 Bug 840: fix log_defer_output option of pipe transport
783 JH/11 Bug 830: use same host for all RCPTS of a message, even under
784 hosts_randomize. This matters a lot when combined with mua_wrapper.
786 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
787 ${quote_pgsql:<string>} operator.
789 JH/13 Bug 1708: avoid misaligned access in cached lookup.
791 JH/14 Change header file name for freeradius-client. Relevant if compiling
792 with Radius support; from the Gentoo tree and checked under Fedora.
794 JH/15 Bug 1712: Introduce $prdr_requested flag variable
796 JH/16 Bug 1714: Permit an empty string as expansion result for transport
797 option transport_filter, meaning no filtering.
799 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
801 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
802 defaults to "*" (all hosts). The variable is now available when not built
803 with TLS, default unset, mainly to enable keeping the testsuite sane.
804 If a server certificate is not supplied (via tls_certificate) an error is
805 logged, and clients will find TLS connections fail on startup. Presumably
806 they will retry in-clear.
807 Packagers of Exim are strongly encouraged to create a server certificate
808 at installation time.
810 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
811 with the $config_file variable.
813 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
814 in transport context, after the attempt, and per-recipient. The latter type
815 is per host attempted. The event data is the error message, and the errno
816 information encodes the lookup type (A vs. MX) used for the (first) host,
817 and the trailing two digits of the smtp 4xx response.
819 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
820 to write to mainlog (or rejectlog, paniclog) in the window between file
821 creation and permissions/ownership being changed. Particularly affects
822 installations where exicyclog is run as root, rather than exim user;
823 result is that the running daemon panics and dies.
825 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
827 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
828 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
829 "pri" and "weight". Note that the previous implicit priority given by the
830 list order is no longer honoured.
832 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
835 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
836 by defining SUPPORT_SOCKS.
838 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
839 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
840 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
841 variables are renamed to proxy_{local,external}_{address,port}.
843 JH/27 Move Internationalisation support from Experimental to mainline, enabled
844 for a build by defining SUPPORT_I18N
846 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
847 of the query string, and make ${quote_redis:} do that quoting.
849 JH/29 Move Events support from Experimental to mainline, enabled by default
850 and removable for a build by defining DISABLE_EVENT.
852 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
854 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
855 cached by the daemon.
857 JH/32 Move Redis support from Experimental to mainline, enabled for a build
858 by defining LOOKUP_REDIS. The libhiredis library is required.
860 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
861 keys are given for lookup.
863 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
864 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
865 only supported when built with TLS support. The PolarSSL SHA routines
866 are still used when the TLS library is too old for convenient support.
868 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
869 openssl_options), for security. OpenSSL forces this from version 1.1.0
870 server-side so match that on older versions.
872 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
873 allocation for $value could be released as the expansion processing
874 concluded, but leaving the global pointer active for it.
876 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
877 and to use the domains and local_parts ACL conditions.
879 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
880 incorrectly not doubled on cutthrough transmission, hence seen as a
881 body-termination at the receiving system - resulting in truncated mails.
882 Commonly the sender saw a TCP-level error, and retransmitted the message
883 via the normal store-and-forward channel. This could result in duplicates
884 received - but deduplicating mailstores were liable to retain only the
885 initial truncated version.
887 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
889 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
891 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
892 we're in there, support oversigning also; bug 1309.
894 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
896 HS/04 Add support for keep_environment and add_environment options.
898 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
899 either intentional arithmetic overflow during PRNG, or testing config-
902 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
903 delivery resulted in actual delivery. Cancel cutthrough before DATA
906 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
907 rejects a recipient: pass the reject to the originator.
909 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
910 Many were false-positives and ignorable, but it's worth fixing the
913 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
914 for the new environment-manipulation done at startup. Move the routines
915 from being local to tls.c to being global via the os.c file.
917 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
918 an extract embedded as result-arg for a map, the first arg for extract
919 is unavailable so we cannot tell if this is a numbered or keyed
920 extraction. Accept either.
926 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
929 JH/02 The smtp transport option "multi_domain" is now expanded.
931 JH/03 The smtp transport now requests PRDR by default, if the server offers
934 JH/04 Certificate name checking on server certificates, when exim is a client,
935 is now done by default. The transport option tls_verify_cert_hostnames
936 can be used to disable this per-host. The build option
937 EXPERIMENTAL_CERTNAMES is withdrawn.
939 JH/05 The value of the tls_verify_certificates smtp transport and main options
940 default to the word "system" to access the system default CA bundle.
941 For GnuTLS, only version 3.0.20 or later.
943 JH/06 Verification of the server certificate for a TLS connection is now tried
944 (but not required) by default. The verification status is now logged by
945 default, for both outbound TLS and client-certificate supplying inbound
948 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
951 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
952 Status Notification (bounce) messages are now MIME format per RFC 3464.
953 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
954 under the control of the dsn_advertise_hosts option, and routers may
955 have a dsn_lasthop option.
957 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
958 default, modifiable by a malware= option. The list separator for
959 the options can now be changed in the usual way. Bug 68.
961 JH/10 The smtp_receive_timeout main option is now expanded before use.
963 JH/11 The incoming_interface log option now also enables logging of the
964 local interface on delivery outgoing connections.
966 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
967 if the interface and destination host and port all match.
969 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
972 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
973 Patch from Andrew Lewis.
975 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
976 now supports optional time-restrictions, weighting, and priority
977 modifiers per server. Patch originally by <rommer@active.by>.
979 JH/16 The spamd_address main option now supports a mixed list of local
980 and remote servers. Remote servers can be IPv6 addresses, and
981 specify a port-range.
983 JH/17 Bug 68: The spamd_address main option now supports an optional
984 timeout value per server.
986 JH/18 Bug 1581: Router and transport options headers_add/remove can
987 now have the list separator specified.
989 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
992 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
995 JH/21 Support for the A6 type of dns record is withdrawn.
997 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
998 rather than the verbs used.
1000 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1001 from 255 to 1024 chars.
1003 JH/24 Verification callouts now attempt to use TLS by default.
1005 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1006 are generic router options now. The defaults didn't change.
1008 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1009 Original patch from Alexander Shikoff, worked over by JH.
1011 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1012 files if ZCAT_COMMAND is not executable.
1014 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1016 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1018 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1019 Normally benign, it bites when the pair was led to by a CNAME;
1020 modern usage is to not canonicalize the domain to a CNAME target
1021 (and we were inconsistent anyway for A-only vs AAAA+A).
1023 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1025 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1026 when evaluating $sender_host_dnssec.
1028 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1029 $sender_helo_dnssec variable.
1031 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1033 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1035 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1037 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1038 documented as working, but never had. Support all but $spam_report.
1040 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1041 added for tls authenticator.
1043 HS/03 Add perl_taintmode main config option
1049 TL/01 When running the test suite, the README says that variables such as
1050 no_msglog_check are global and can be placed anywhere in a specific
1051 test's script, however it was observed that placement needed to be near
1052 the beginning for it to behave that way. Changed the runtest perl
1053 script to read through the entire script once to detect and set these
1054 variables, reset to the beginning of the script, and then run through
1055 the script parsing/test process like normal.
1057 TL/02 The BSD's have an arc4random API. One of the functions to induce
1058 adding randomness was arc4random_stir(), but it has been removed in
1059 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1060 function when detected.
1062 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1063 cause callback expansion.
1065 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1066 syntax errors in an expansion can be treated as a string instead of
1067 logging or causing an error, due to the internal use of bool_lax
1068 instead of bool when processing it.
1070 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1071 server certificates when making smtp deliveries.
1073 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1075 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1077 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1079 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1080 Merged patch from Sebastian Wiedenroth.
1082 JH/05 Fix results-pipe from transport process. Several recipients, combined
1083 with certificate use, exposed issues where response data items split
1084 over buffer boundaries were not parsed properly. This eventually
1085 resulted in duplicates being sent. This issue only became common enough
1086 to notice due to the introduction of connection certificate information,
1087 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1089 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1090 size buffer was used, resulting in syntax errors when an expansion
1093 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1094 version 3.3.6 or later.
1096 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1097 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1098 both become "event_action", the variables become $event_name, $event_data
1099 and $event_defer_errno. There is a new variable $verify_mode, usable in
1100 routers, transports and related events. The tls:cert event is now also
1101 raised for inbound connections, if the main configuration event_action
1104 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1105 early OCSP support, but no stapling (appears to be less than 1.0.0).
1107 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1108 server certificate names available under the smtp transport option
1109 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1112 JH/10 Time-related extraction expansions from certificates now use the main
1113 option "timezone" setting for output formatting, and are consistent
1114 between OpenSSL and GnuTLS compilations. Bug 1541.
1116 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1117 encoded parameter in the incoming message. Bug 1558.
1119 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1120 include certificate info, eximon was claiming there were spoolfile
1123 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1125 JH/14 Log delivery-related information more consistently, using the sequence
1126 "H=<name> [<ip>]" wherever possible.
1128 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1129 are problematic for Debian distribution, omit them from the release
1132 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1134 JH/16 Fix string representation of time values on 64bit time_t architectures.
1137 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1138 output list separator was used.
1143 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1144 checkers that were complaining about end of non-void function with no
1147 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1148 This was a regression introduced in 4.83 by another bugfix.
1150 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1152 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1153 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1159 TF/01 Correctly close the server side of TLS when forking for delivery.
1161 When a message was received over SMTP with TLS, Exim failed to clear up
1162 the incoming connection properly after forking off the child process to
1163 deliver the message. In some situations the subsequent outgoing
1164 delivery connection happened to have the same fd number as the incoming
1165 connection previously had. Exim would try to use TLS and fail, logging
1166 a "Bad file descriptor" error.
1168 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1169 utilities have not been installed.
1171 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1172 temporary space as the ACL may create new global variables.
1174 TL/01 LDAP support uses per connection or global context settings, depending
1175 upon the detected version of the libraries at build time.
1177 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1178 to extract and use the src ip:port in logging and expansions as if it
1179 were a direct connection from the outside internet. PPv2 support was
1180 updated based on HAProxy spec change in May 2014.
1182 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1184 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1185 Properly escape header and check for NULL return.
1187 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1190 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1192 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1193 characters in header names, implemented as a verify condition.
1194 Contributed by Michael Fischer v. Mollard.
1196 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1197 results permerror and temperror. Previous values are deprecated but
1198 still accepted. In a future release, err_perm and err_temp will be
1199 completely removed, which will be a backward incompatibility if the
1200 ACL tests for either of these two old results. Patch contributed by
1201 user bes-internal on the mailing list.
1203 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1205 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1206 selectors, in both main and reject logs.
1208 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1211 JH/07 Add malware type "sock" for talking to simple daemon.
1213 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1215 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1216 routers/transports under cutthrough routing.
1218 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1219 numbers. Touch up "bool" conditional to keep the same definition.
1221 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1223 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1225 JH/12 Expand items in router/transport headers_add or headers_remove lists
1226 individually rather than the list as a whole. Bug 1452.
1228 Required for reasonable handling of multiple headers_ options when
1229 they may be empty; requires that headers_remove items with embedded
1230 colons must have them doubled (or the list-separator changed).
1232 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1233 view the policy declared in the DMARC record. Currently, $dmarc_status
1234 is a combined value of both the record presence and the result of the
1237 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1239 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1240 dnslookup router and the smtp transport (applying to the forward
1243 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1244 of ldap servers used for a specific lookup. Patch provided by Heiko
1247 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1248 New variable $lookup_dnssec_authenticated for observability.
1250 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1251 Patch submitted by Lars Timman.
1253 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1255 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1256 Requires trusted mode and valid format message id, aborts otherwise.
1257 Patch contributed by Heiko Schlichting.
1259 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1260 certextract with support for various fields. Bug 1358.
1262 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1263 is requested by default, modifiable by smtp transport option
1266 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1267 operate on certificate variables to give certificate fingerprints
1268 Also new ${sha256:cert_variable}.
1270 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1272 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1273 Christian Aistleitner.
1275 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1277 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1278 file. Patch from Wolfgang Breyha.
1280 JH/25 Expand the coverage of the delivery $host and $host_address to
1281 client authenticators run in verify callout. Bug 1476.
1283 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1284 align with daemon_smtp_ports. Bug 72.
1286 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1287 support and error reporting did not work properly.
1289 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1290 and is readable. Patch from Andrew Colin Kissa.
1292 TL/14 Enhance documentation of ${run expansion and how it parses the
1293 commandline after expansion, particularly in the case when an
1294 unquoted variable expansion results in an empty value.
1296 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1298 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1299 by renaming away from it. Addresses GH issue 15, reported by
1302 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1304 TL/15 SECURITY: prevent double expansion in math comparison functions
1305 (can expand unsanitized data). Not remotely exploitable.
1312 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1314 PP/02 Make -n do something, by making it not do something.
1315 When combined with -bP, the name of an option is not output.
1317 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1320 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1321 $sender_host_name and config options to manage this, and basic check
1324 PP/05 DSCP support for outbound connections and control modifier for inbound.
1326 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1327 (Only plugin which currently uses this is kerberos4, which nobody should
1328 be using, but we should make it available and other future plugins might
1329 conceivably use it, even though it would break NAT; stuff *should* be
1330 using channel bindings instead).
1332 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1333 name; added for Sendmail compatibility; requires admin caller.
1334 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1335 just ignore it); requires trusted caller.
1336 Also parse but ignore: -Ac -Am -X<logfile>
1339 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1341 TL/02 Add +smtp_confirmation as a default logging option.
1343 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1344 Patch by Magnus Holmgren from 2007-02-20.
1346 TL/04 Bugzilla 1281 - Spec typo.
1347 Bugzilla 1283 - Spec typo.
1348 Bugzilla 1290 - Spec grammar fixes.
1350 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1352 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1354 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1355 mailing list by by Dmitry Isaikin.
1357 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1359 JH/02 Support "G" suffix to numbers in ${if comparisons.
1361 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1363 NM/01 Bugzilla 1197 - Spec typo
1364 Bugzilla 1196 - Spec examples corrections
1366 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1368 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1369 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1372 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1373 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1375 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1376 "acl {{name}{arg}...}", and optional args on acl condition
1379 JH/05 Permit multiple router/transport headers_add/remove lines.
1381 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1383 JH/07 Avoid using a waiting database for a single-message-only transport.
1384 Performance patch from Paul Fisher. Bugzilla 1262.
1386 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1389 JH/09 Add $headers_added variable, with content from use of ACL modifier
1390 add_header (but not yet added to the message). Bugzilla 199.
1392 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1393 Pulled from Bugzilla 817 by Wolfgang Breyha.
1395 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1397 (nb: this is the same fix as in Exim 4.80.1)
1399 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1402 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1405 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1407 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1408 advertise SMTP AUTH mechanism to us, instead of a generic
1409 protocol violation error. Also, make Exim more robust to bad
1410 data from the Dovecot auth socket.
1412 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1414 When a queue runner is handling a message, Exim first routes the
1415 recipient addresses, during which it prunes them based on the retry
1416 hints database. After that it attempts to deliver the message to
1417 any remaining recipients. It then updates the hints database using
1420 So if a recipient address works intermittently, it can get repeatedly
1421 deferred at routing time. The retry hints record remains fresh so the
1422 address never reaches the final cutoff time.
1424 This is a fairly common occurrence when a user is bumping up against
1425 their storage quota. Exim had some logic in its local delivery code
1426 to deal with this. However it did not apply to per-recipient defers
1427 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1429 This change adds a proper retry rule check during routing so that the
1430 final cutoff time is checked against the message's age. We only do
1431 this check if there is an address retry record and there is not a
1432 domain retry record; this implies that previous attempts to handle
1433 the address had the retry_use_local_parts option turned on. We use
1434 this as an approximation for the destination being like a local
1435 delivery, as in LMTP.
1437 I suspect this new check makes the old local delivery cutoff check
1438 redundant, but I have not verified this so I left the code in place.
1440 TF/02 Correct gecos expansion when From: is a prefix of the username.
1442 Test 0254 submits a message to Exim with the header
1446 When I ran the test suite under the user fanf2, Exim expanded
1447 the header to contain my full name, whereas it should have added
1448 a Resent-Sender: header. It erroneously treats any prefix of the
1449 username as equal to the username.
1451 This change corrects that bug.
1453 GF/01 DCC debug and logging tidyup
1454 Error conditions log to paniclog rather than rejectlog.
1455 Debug lines prefixed by "DCC: " to remove any ambiguity.
1457 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1459 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1460 Bug spotted by Jeremy Harris; was flawed since initial commit.
1461 Would have resulted in OCSP responses post-SNI triggering an Exim
1462 NULL dereference and crash.
1464 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1466 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1467 Bug detection, analysis and fix by Samuel Thibault.
1468 Bugzilla 1331, Debian bug #698092.
1470 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1472 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1473 Server implementation by Todd Lyons, client by JH.
1474 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1475 config variable "prdr_enable" controls whether the server
1476 advertises the facility. If the client requests PRDR a new
1477 acl_data_smtp_prdr ACL is called once for each recipient, after
1478 the body content is received and before the acl_smtp_data ACL.
1479 The client is controlled by both of: a hosts_try_prdr option
1480 on the smtp transport, and the server advertisement.
1481 Default client logging of deliveries and rejections involving
1482 PRDR are flagged with the string "PRDR".
1484 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1485 fclose(). Diagnosis by Todd Lyons.
1487 PP/17 Update configure.default to handle IPv6 localhost better.
1488 Patch by Alain Williams (plus minor tweaks).
1491 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1492 This is now consistent with GnuTLS, and is now documented: the
1493 previous undocumented portable approach to treating the option as
1494 unset was to force an expansion failure. That still works, and
1495 an empty string is now equivalent.
1497 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1498 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1499 not performing validation itself.
1501 PP/20 Added force_command boolean option to pipe transport.
1502 Patch from Nick Koston, of cPanel Inc.
1504 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1507 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1509 PP/21 Fix eximon continuous updating with timestamped log-files.
1510 Broken in a format-string cleanup in 4.80, missed when I repaired the
1511 other false fix of the same issue.
1512 Report and fix from Heiko Schlichting.
1515 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1516 Report from Prashanth Katuri.
1518 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1519 It's SecureTransport, so affects any MacOS clients which use the
1520 system-integrated TLS libraries, including email clients.
1522 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1523 using a MIME ACL for non-SMTP local injection.
1524 Report and assistance in diagnosis by Warren Baker.
1526 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1528 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1530 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1531 last id that failed so it may be referenced in subsequent ACL's.
1533 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1536 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1537 ldap library initialization, allowing self-signed CA's to be
1538 used. Also properly sets require_cert option later in code by
1539 using NULL (global ldap config) instead of ldap handle (per
1540 session). Bug diagnosis and testing by alxgomz.
1542 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1543 the src/util/ subdirectory.
1545 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1546 renamed to Transport Post Delivery Action by Jeremy Harris, as
1549 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1550 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1551 redis_servers = needs to be configured which will be used by the redis
1552 lookup. Patch from Warren Baker, of The Packet Hub.
1554 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1556 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1557 hostname or reverse DNS when processing a host list. Used suggestions
1558 from multiple comments on this bug.
1560 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1562 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1563 Missed a few lines, added it to make the runtest require no keyboard
1566 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1567 contains upper case chars. Make router use caseful_local_part.
1569 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1570 support when GnuTLS has been built with p11-kit.
1576 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1578 This, or similar/improved, will also be change PP/11 of 4.82.
1584 PP/01 Handle short writes when writing local log-files.
1585 In practice, only affects FreeBSD (8 onwards).
1586 Bugzilla 1053, with thanks to Dmitry Isaikin.
1588 NM/01 Bugzilla 949 - Documentation tweak
1590 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1593 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1595 PP/02 Implemented gsasl authenticator.
1597 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1599 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1600 `pkg-config foo` for cflags/libs.
1602 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1603 with rest of GSASL and with heimdal_gssapi.
1605 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1606 `pkg-config foo` for cflags/libs for the TLS implementation.
1608 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1609 properties get this fed in as external SSF. A number of robustness
1610 and debugging improvements to the cyrus_sasl authenticator.
1612 PP/08 cyrus_sasl server now expands the server_realm option.
1614 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1615 Patch by Jeremy Harris.
1617 PP/10 Added dbmjz lookup type.
1619 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1621 PP/12 MAIL args handles TAB as well as SP, for better interop with
1622 non-compliant senders.
1623 Analysis and variant patch by Todd Lyons.
1625 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1626 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1627 Patch from Dirk Mueller <dmueller@suse.com>
1629 PP/13 tls_peerdn now print-escaped for spool files.
1630 Observed some $tls_peerdn in wild which contained \n, which resulted
1631 in spool file corruption.
1633 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1634 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1635 or write after TLS renegotiation, which otherwise led to messages
1638 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1639 as a tracking header (ie: a signed header comes before the signature).
1640 Patch from Wolfgang Breyha.
1642 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1643 comma-sep list; embedded commas doubled.
1645 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1647 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1649 Report and patch from Dmitry Banschikov.
1651 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1652 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1653 failures. We appear to now support TLS1.1+ with Exim.
1655 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1656 lets Exim select keys and certificates based upon TLS SNI from client.
1657 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1658 before an outbound SMTP session. New log_selector, +tls_sni.
1660 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1661 NULL dereference. Report and patch from Alun Jones.
1663 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1664 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1665 Not seeing resolver debug output on NetBSD, but suspect this is a
1666 resolver implementation change.
1668 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1669 Left warnings. Added "eximon gdb" invocation mode.
1671 PP/21 Defaulting "accept_8bitmime" to true, not false.
1673 PP/22 Added -bw for inetd wait mode support.
1675 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1676 locate the relevant includes and libraries. Made this the default.
1678 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1679 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1681 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1682 This may cause build issues on older platforms.
1684 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1685 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1686 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1687 Added SNI support via GnuTLS too.
1688 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1690 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1692 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1693 Applied second patch from Janne, implementing suggestion to default
1694 multiple-strings-in-record handling to match SPF spec.
1696 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1698 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1699 read-only, out of scope).
1700 Patch from Wolfgang Breyha, report from Stuart Northfield.
1702 PP/29 Fix three issues highlighted by clang analyser static analysis.
1703 Only crash-plausible issue would require the Cambridge-specific
1704 iplookup router and a misconfiguration.
1705 Report from Marcin Mirosław.
1707 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1709 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1710 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1711 As part of this, removing so much warning spew let me fix some minor
1712 real issues in debug logging.
1714 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1715 assignment on my part. Fixed.
1717 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1718 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1719 Janne Snabb (who went above and beyond: thank you).
1721 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1722 string otherwise requires a connection and a bunch more work and it's
1723 relatively easy to get wrong. Should also expose TLS library linkage
1726 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1727 64-bit ${eval} (JH/03).
1729 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1730 GNU libc to support some of the 64-bit stuff, should not lead to
1731 conflicts. Defined before os.h is pulled in, so if a given platform
1732 needs to override this, it can.
1734 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1735 protection layer was required, which is not implemented.
1736 Bugzilla 1254, patch from Wolfgang Breyha.
1738 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1739 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1740 tls_dhparam take prime identifiers. Also unbreak combination of
1741 OpenSSL+DH_params+TLSSNI.
1743 PP/39 Disable SSLv2 by default in OpenSSL support.
1749 PP/01 Solaris build fix for Oracle's LDAP libraries.
1750 Bugzilla 1109, patch from Stephen Usher.
1752 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1754 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1757 TF/02 Fix a couple more cases where we did not log the error message
1758 when unlink() failed. See also change 4.74-TF/03.
1760 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1761 lock up or crash if it happened to be inside a call to libc when it
1762 got a SIGUSR1 from exiwhat.
1764 The SIGUSR1 handler appends the current process status to the process
1765 log which is later printed by exiwhat. It used to use the general
1766 purpose logging code to do this, but several functions it calls are
1767 not safe for signals.
1769 The new output code in the SIGUSR1 handler is specific to the process
1770 log, and simple enough that it's easy to inspect for signal safety.
1771 Removing some special cases also simplifies the general logging code.
1772 Removing the spurious timestamps from the process log simplifies
1775 TF/04 Improved ratelimit ACL condition.
1777 The /noupdate option has been deprecated in favour of /readonly which
1778 has clearer semantics. The /leaky, /strict, and /readonly update modes
1779 are mutually exclusive. The update mode is no longer included in the
1780 database key; it just determines when the database is updated. (This
1781 means that when you upgrade Exim will forget old rate measurements.)
1783 Exim now checks that the per_* options are used with an update mode that
1784 makes sense for the current ACL. For example, when Exim is processing a
1785 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1786 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1787 must specify per_mail/readonly. If you omit the update mode it defaults to
1788 /leaky where that makes sense (as before) or /readonly where required.
1790 The /noupdate option is now undocumented but still supported for
1791 backwards compatibility. It is equivalent to /readonly except that in
1792 ACLs where /readonly is required you may specify /leaky/noupdate or
1793 /strict/noupdate which are treated the same as /readonly.
1795 A useful new feature is the /count= option. This is a generalization
1796 of the per_byte option, so that you can measure the throughput of other
1797 aggregate values. For example, the per_byte option is now equivalent
1798 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1800 The per_rcpt option has been generalized using the /count= mechanism
1801 (though it's more complicated than the per_byte equivalence). When it is
1802 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1803 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1804 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1805 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1806 Note that using per_rcpt with a non-readonly update mode in more than
1807 one ACL will cause the recipients to be double-counted. (The per_mail
1808 and per_byte options don't have this problem.)
1810 The handling of very low rates has changed slightly. If the computed rate
1811 is less than the event's count (usually one) then this event is the first
1812 after a long gap. In this case the rate is set to the same as this event's
1813 count, so that the first message of a spam run is counted properly.
1815 The major new feature is a mechanism for counting the rate of unique
1816 events. The new per_addr option counts the number of different
1817 recipients that someone has sent messages to in the last time period. It
1818 behaves like per_rcpt if all the recipient addresses are different, but
1819 duplicate recipient addresses do not increase the measured rate. Like
1820 the /count= option this is a general mechanism, so the per_addr option
1821 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1822 example, measure the rate that a client uses different sender addresses
1823 with the options per_mail/unique=$sender_address. There are further
1824 details in the main documentation.
1826 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1828 TF/06 Removed a few PCRE remnants.
1830 TF/07 Automatically extract Exim's version number from tags in the git
1831 repository when doing development or release builds.
1833 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1834 Bugzilla 879. Patch from Paul Fisher.
1836 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1837 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1840 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1842 PP/05 Variable $av_failed, true if the AV scanner deferred.
1843 Bugzilla 1078. Patch from John Horne.
1845 PP/06 Stop make process more reliably on build failure.
1846 Bugzilla 1087. Patch from Heiko Schlittermann.
1848 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1849 Bugzilla 1089. Patch from Heiko Schlittermann.
1851 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1852 Bugzilla 1131. Patch from Holger Weiß.
1854 PP/09 Handle IPv6 addresses with SPF.
1855 Bugzilla 860. Patch from Wolfgang Breyha.
1857 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1859 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1862 PP/11 match_* no longer expand right-hand-side by default.
1863 New compile-time build option, EXPAND_LISTMATCH_RHS.
1864 New expansion conditions, "inlist", "inlisti".
1866 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1868 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1870 PP/14 fix log_write() format string regression from TF/03.
1871 Bugzilla 1152. Patch from Dmitry Isaikin.
1877 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1879 PP/02 Harmonised TLS library version reporting; only show if debugging.
1880 Layout now matches that introduced for other libraries in 4.74 PP/03.
1882 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1884 PP/04 New "dns_use_edns0" global option.
1886 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1889 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1890 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1892 TK/01 Updated PolarSSL code to 0.14.2.
1893 Bugzilla 1097. Patch from Andreas Metzler.
1895 PP/07 Catch divide-by-zero in ${eval:...}.
1896 Fixes bugzilla 1102.
1898 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1901 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1902 format-string attack -- SECURITY: remote arbitrary code execution.
1904 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1905 time unintentionally subject to list matching rules, letting the header
1906 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1907 arbitrary string expansion). This allowed for information disclosure.
1909 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1910 INT_MIN/-1 -- value coerced to INT_MAX.
1916 NM/01 Workaround for PCRE version dependency in version reporting
1919 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1920 This fixes portability to compilers other than gcc, notably
1921 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1923 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1924 makefiles for portability to HP-UX and POSIX correctness.
1926 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1927 Also via indented variable definition in the Makefile.
1928 (Debugging by Oliver Heesakkers).
1930 PP/02 Restore caching of spamd results with expanded spamd_address.
1931 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1933 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1934 Improves build reliability. Fix from: Frank Elsner
1936 NM/02 Fix wide character breakage in the rfc2047 coding
1937 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1939 NM/03 Allow underscore in dnslist lookups
1940 Fixes bug 1026. Patch from Graeme Fowler
1942 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1943 Code patches from Adam Ciarcinski of NetBSD.
1945 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1948 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1949 is logged, to avoid truncation. Patch from John Horne.
1951 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1952 Patch from Jakob Hirsch.
1954 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1955 SQL string expansion failure details.
1956 Patch from Andrey Oktyabrski.
1958 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1959 Patch from Simon Arlott.
1961 PP/09 New lookups functionality failed to compile on old gcc which rejects
1962 extern declarations in function scope.
1963 Patch from Oliver Fleischmann
1965 PP/10 Use sig_atomic_t for flags set from signal handlers.
1966 Check getgroups() return and improve debugging.
1967 Fixed developed for diagnosis in bug 927 (which turned out to be
1970 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1971 Patch from Mark Zealey.
1973 PP/12 Bugzilla 1056: Improved spamd server selection.
1974 Patch from Mark Zealey.
1976 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1977 Based on patch from Heiko Schlittermann.
1979 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1980 Patch from Uwe Doering, sign-off by Michael Haardt.
1982 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1983 variable declaration deep within a block. Bug and patch from
1986 PP/15 lookups-Makefile IRIX compatibility coercion.
1988 PP/16 Make DISABLE_DKIM build knob functional.
1990 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1991 Patch by Simon Arlott
1993 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1994 variable argument macros. Our copy now differs from upstream.
2000 TF/01 Failure to get a lock on a hints database can have serious
2001 consequences so log it to the panic log.
2003 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2004 controlled using the smtp_confirmation log selector.
2006 TF/03 Include the error message when we fail to unlink a spool file.
2008 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2009 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2010 for maintaining out-of-tree patches for some time.
2012 PP/01 Bugzilla 139: Documentation and portability issues.
2013 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2014 Handle per-OS dynamic-module compilation flags.
2016 PP/02 Let /dev/null have normal permissions.
2017 The 4.73 fixes were a little too stringent and complained about the
2018 permissions on /dev/null. Exempt it from some checks.
2019 Reported by Andreas M. Kirchwitz.
2021 PP/03 Report version information for many libraries, including
2022 Exim version information for dynamically loaded libraries. Created
2023 version.h, now support a version extension string for distributors
2024 who patch heavily. Dynamic module ABI change.
2026 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2027 privilege escalation vulnerability whereby the Exim run-time user
2028 can cause root to append content of the attacker's choosing to
2031 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2034 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2035 If dropping privileges for untrusted macros, we disabled normal logging
2036 on the basis that it would fail; for the Exim run-time user, this is not
2037 the case, and it resulted in successful deliveries going unlogged.
2038 Fixed. Reported by Andreas Metzler.
2044 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2045 only prepend for the Resent-* case. Fixes regression introduced in
2046 Exim 4.70 by NM/22 for Bugzilla 607.
2048 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2049 increasing numbers of administrators be bitten by this.
2051 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2053 PP/03 Bugzilla 994: added openssl_options main configuration option.
2055 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2057 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2059 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2061 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2062 without a peer certificate, leading to a segfault because of an
2063 assumption that peers always have certificates. Be a little more
2064 paranoid. Problem reported by Martin Tscholak.
2066 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2067 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2068 NB: ClamAV planning to remove STREAM in "middle of 2010".
2069 CL also introduces -bmalware, various -d+acl logging additions and
2070 more caution in buffer sizes.
2072 PP/09 Implemented reverse_ip expansion operator.
2074 PP/10 Bugzilla 937: provide a "debug" ACL control.
2076 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2078 PP/12 Bugzilla 973: Implement --version.
2080 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2082 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2084 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2086 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2087 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2088 ignore trailing whitespace.
2090 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2092 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2095 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2096 Notification from Dr Andrew Aitchison.
2098 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2099 ExtendedDetectionInfo response format.
2100 Notification from John Horne.
2102 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2105 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2106 XSL and documented dependency on system catalogs, with examples of how
2109 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2112 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2113 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2114 configuration file which is writeable by the Exim user or group.
2116 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2117 of configuration files to cover files specified with the -C option if
2118 they are going to be used with root privileges, not just the default
2121 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2122 option (effectively making it always true).
2124 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2125 files to be used while preserving root privileges.
2127 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2128 that rogue child processes cannot use them.
2130 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2131 run-time user, instead of root.
2133 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2134 Exim run-time user without dropping privileges.
2136 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2137 result string, instead of calling string_vformat() twice with the same
2140 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2141 for other users. Others should always drop root privileges if they use
2142 -C on the command line, even for a whitelisted configure file.
2144 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2146 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2152 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2153 $data_path, and $header_path variables; fixed documentation bugs and
2156 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2157 exipick to access non-standard spools, including the "frozen" queue
2160 NM/01 Bugzilla 965: Support mysql stored procedures.
2161 Patch from Alain Williams
2163 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2165 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2166 Patch from Andreas Metzler
2168 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2169 Patch from Kirill Miazine
2171 NM/05 Bugzilla 671: Added umask to procmail example.
2173 JJ/03 installed exipick 20100323.0, fixing doc bug
2175 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2176 directory. Notification and patch from Dan Rosenberg.
2178 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2180 TK/02 Improve log output when DKIM signing operation fails.
2182 MH/01 Treat the transport option dkim_domain as a colon separated
2183 list, not as a single string, and sign the message with each element,
2184 omitting multiple occurences of the same signer.
2186 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2187 Bugzilla 985, 986. Patch by Simon Arlott
2189 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2190 Patch by Simon Arlott
2192 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2193 MBX locking. Notification from Dan Rosenberg.
2199 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2201 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2203 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2205 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2207 NM/04 Bugzilla 915: Flag broken perl installation during build.
2213 TK/01 Added patch by Johannes Berg that expands the main option
2214 "spamd_address" if it starts with a dollar sign.
2216 TK/02 Write list of recipients to X-Envelope-Sender header when building
2217 the mbox-format spool file for content scanning (suggested by Jakob
2220 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2221 (http://www.dcc-servers.net/) support via dccifd. Activated by
2222 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2224 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2225 by Mark Daniel Reidel <mr@df.eu>.
2227 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2228 When building exim an external PCRE library is now needed -
2229 PCRE is a system library on the majority of modern systems.
2230 See entry on PCRE_LIBS in EDITME file.
2232 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2233 conversation. Added nologin parameter to request.
2234 Patch contributed by Kirill Miazine.
2236 TF/01 Do not log submission mode rewrites if they do not change the address.
2238 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2240 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2241 log files in place. Contributed by Roberto Lima.
2243 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2245 TF/03 Bugzilla 615: When checking the local_parts router precondition
2246 after a local_part_suffix or local_part_prefix option, Exim now
2247 does not use the address's named list lookup cache, since this
2248 contains cached lookups for the whole local part.
2250 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2251 Robert Millan. Documentation is in experimental-spec.txt.
2253 TF/04 Bugzilla 668: Fix parallel build (make -j).
2255 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2257 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2258 Patch provided by Jan Srzednicki.
2260 TF/05 Leading white space used to be stripped from $spam_report which
2261 wrecked the formatting. Now it is preserved.
2263 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2264 that they are available at delivery time.
2266 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2268 TF/08 TLS error reporting now respects the incoming_interface and
2269 incoming_port log selectors.
2271 TF/09 Produce a more useful error message if an SMTP transport's hosts
2272 setting expands to an empty string.
2274 NM/06 Bugzilla 744: EXPN did not work under TLS.
2275 Patch provided by Phil Pennock.
2277 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2278 Patch provided by Richard Godbee.
2280 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2281 acl_smtp_notquit, added index entry.
2283 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2284 Patch provided by Eugene Bujak.
2286 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2287 accept(). Patch provided by Maxim Dounin.
2289 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2290 Patch provided by Phil Pennock.
2292 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2294 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2295 Patch provided by Brad "anomie" Jorsch.
2297 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2298 Patch provided by Dean Brooks.
2300 TK/05 Add native DKIM support (does not depend on external libraries).
2302 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2303 Patch provided by Graeme Fowler.
2305 NM/16 Bugzilla 851: Documentation example syntax fix.
2307 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2309 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2312 NM/19 Bugzilla 745: TLS version reporting.
2313 Patch provided by Phil Pennock.
2315 NM/20 Bugzilla 167: bool: condition support.
2316 Patch provided by Phil Pennock.
2318 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2319 clients. Patch provided by Phil Pennock.
2321 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2322 Patch provided by Brad "anomie" Jorsch.
2324 NM/23 Bugzilla 687: Fix misparses in eximstats.
2325 Patch provided by Heiko Schlittermann.
2327 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2328 Patch provided by Heiko Schlittermann.
2330 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2331 plus update to original patch.
2333 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2335 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2336 Patch provided by David Brownlee.
2338 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2340 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2342 NM/30 Bugzilla 888: TLS documentation bugfixes.
2344 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2346 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2347 Unlike the original bugzilla I have changed all shell scripts in src tree.
2349 NM/33 Bugzilla 898: Transport filter timeout fix.
2350 Patch by Todd Rinaldo.
2352 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2353 Patch by Serge Demonchaux.
2355 NM/35 Bugzilla 39: Base64 decode bug fixes.
2356 Patch by Jakob Hirsch.
2358 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2360 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2362 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2364 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2370 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2371 ALT-N's libdkim that I have put here:
2372 http://duncanthrax.net/exim-experimental/
2374 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2375 were called 'true' and it seems that C99 defines that as a reserved
2376 keyword to be used with 'bool' variable types. That means you could
2377 not include C99-style headers which use bools without triggering
2378 build errors in sieve.c.
2380 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2381 as mailq or other aliases. Changed the --help handling significantly
2382 to do whats expected. exim_usage() emits usage/help information.
2384 SC/01 Added the -bylocaldomain option to eximstats.
2386 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2388 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2390 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2396 PH/01 Another patch from the Sieve maintainer.
2398 PH/02 When an IPv6 address is converted to a string for single-key lookup
2399 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2400 dots are used instead of colons so that keys in lsearch files need not
2401 contain colons. This was done some time before quoting was made available
2402 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2403 (notated using the quote facility) so as to distinguish them from IPv4
2404 keys. This meant that lookups for IP addresses in host lists did not work
2405 for iplsearch lookups.
2407 This has been fixed by arranging for IPv6 addresses to be expressed with
2408 colons if the lookup type is iplsearch. This is not incompatible, because
2409 previously such lookups could never work.
2411 The situation is now rather anomalous, since one *can* have colons in
2412 ordinary lsearch keys. However, making the change in all cases is
2413 incompatible and would probably break a number of configurations.
2415 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2418 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2419 conversion specification without a maximum field width, thereby enabling
2420 a rogue spamd server to cause a buffer overflow. While nobody in their
2421 right mind would setup Exim to query an untrusted spamd server, an
2422 attacker that gains access to a server running spamd could potentially
2423 exploit this vulnerability to run arbitrary code as the Exim user.
2425 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2426 $primary_hostname instead of what libspf2 thinks the hosts name is.
2428 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2429 a directory entry by the name of the lookup key. Previously, if a
2430 symlink pointed to a non-existing file or a file in a directory that
2431 Exim lacked permissions to read, a lookup for a key matching that
2432 symlink would fail. Now it is enough that a matching directory entry
2433 exists, symlink or not. (Bugzilla 503.)
2435 PH/03 The body_linecount and body_zerocount variables are now exported in the
2438 PH/04 Added the $dnslist_matched variable.
2440 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2441 This means they are set thereafter only if the connection becomes
2444 PH/06 Added the client_condition to authenticators so that some can be skipped
2445 by clients under certain conditions.
2447 PH/07 The error message for a badly-placed control=no_multiline_responses left
2448 "_responses" off the end of the name.
2450 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2452 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2453 (without spaces) instead of just copying the configuration text.
2455 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2457 PH/11 Added $max_received_linelength.
2459 PH/12 Added +ignore_defer and +include_defer to host lists.
2461 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2462 way in which PCRE > 7.0 is built.
2464 PH/14 Implemented queue_only_load_latch.
2466 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2467 MAIL command. The effect was to mangle the value on 64-bit systems.
2469 PH/16 Another patch from the Sieve maintainer.
2471 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2473 PH/18 If a system quota error occurred while trying to create the file for
2474 a maildir delivery, the message "Mailbox is full" was not appended to the
2475 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2476 only to a quota excession during the actual writing of the file.
2478 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2479 characters?) which causes problems in log lines. The DN values are now
2480 passed through string_printing() before being added to log lines.
2482 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2483 and InterBase are left for another time.)
2485 PH/21 Added message_body_newlines option.
2487 PH/22 Guard against possible overflow in moan_check_errorcopy().
2489 PH/23 POSIX allows open() to be a macro; guard against that.
2491 PH/24 If the recipient of an error message contained an @ in the local part
2492 (suitably quoted, of course), incorrect values were put in $domain and
2493 $local_part during the evaluation of errors_copy.
2499 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2500 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2503 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2504 issue a MAIL command.
2506 PH/02 In an ACL statement such as
2508 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2510 if a client was not listed at all, or was listed with a value other than
2511 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2512 the condition was not true (as it should be), so access was not denied.
2513 The bug was that the ! inversion was incorrectly passed on to the second
2514 item. This has been fixed.
2516 PH/03 Added additional dnslists conditions == and =& which are different from
2517 = and & when the dns lookup returns more than one IP address.
2519 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2520 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2522 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2523 FSYNC, which compiles an option called disable_fsync that allows for
2524 bypassing fsync(). The documentation is heavily laced with warnings.
2526 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2528 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2529 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2530 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2531 including adding "make clean"; (3) Added -fPIC when compiling the test
2532 dynamically loaded module, to get rid of a warning.
2534 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2535 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2536 = 0s. The bug is otherwise harmless.
2538 PH/07 There was a bug in the dovecot authenticator such that the value of
2539 $auth1 could be overwritten, and so not correctly preserved, after a
2540 successful authentication. This usually meant that the value preserved by
2541 the server_setid option was incorrect.
2543 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2545 PH/09 Installed PCRE release 7.0.
2547 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2548 run for batched SMTP input. It is now run at the start of every message
2549 in the batch. While fixing this I discovered that the process information
2550 (output by running exiwhat) was not always getting set for -bs and -bS
2551 input. This is fixed, and it now also says "batched" for BSMTP.
2553 PH/11 Added control=no_pipelining.
2555 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2556 patch, slightly modified), and move the expansion of helo_data till after
2557 the connection is made in the smtp transport (so it can use these
2560 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2562 PH/14 Added log_selector = +pid.
2564 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2566 PH/16 Add ${if forany and ${if forall.
2568 PH/17 Added dsn_from option to vary the From: line in DSNs.
2570 PH/18 Flush SMTP output before performing a callout, unless control =
2571 no_callout_flush is set.
2573 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2574 was true (the default) a successful delivery failed to delete the retry
2575 item, thus causing premature timeout of the address. The bug is now
2578 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2580 PH/21 Long custom messages for fakedefer and fakereject are now split up
2581 into multiline responses in the same way that messages for "deny" and
2582 other ACL rejections are.
2584 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2585 with slight modification.
2587 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2588 draft, changing the syntax and factoring some duplicate code".
2590 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2591 for deliveries of the second and subsequent messages over the same SMTP
2594 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2595 ${reduce, with only minor "tidies".
2597 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2599 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2600 expansion side effects.
2602 PH/27 When a message times out after an over-quota error from an Exim-imposed
2603 quota, the bounce message says "mailbox is full". This message was not
2604 being given when it was a system quota that was exceeded. It now should
2607 MH/03 Made $recipients available in local_scan(). local_scan() already has
2608 better access to the recipient list through recipients_list[], but
2609 $recipients can be useful in postmaster-provided expansion strings.
2611 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2612 in the case of a MAIL command with additional options following the
2613 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2614 were accidentally chopped off.
2616 PH/29 SMTP synchronization checks are implemented when a command is read -
2617 there is a check that no more input is waiting when there shouldn't be
2618 any. However, for some commands, a delay in an ACL can mean that it is
2619 some time before the response is written. In this time, more input might
2620 arrive, invalidly. So now there are extra checks after an ACL has run for
2621 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2622 pipelining has not been advertised.
2624 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2626 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2627 returned in the SMTP response when smtp_return_error_details was false.
2628 This has been fixed.
2630 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2631 instead of the C I/O that was originally supplied, because problems were
2632 reported on Solaris.
2634 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2635 Exim which did not show up earlier: it was assuming that a call to
2636 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2637 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2638 was a macro that became an assignment, so it seemed to work. This has
2639 changed to a proper function call with a void return, hence the compile
2640 error. Exim's code has been fixed.
2642 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2645 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2647 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2649 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2650 because it turns out to be too big or there is a timeout) there is no
2651 "Completed" line in the log. When some messages of this type were
2652 selected by exigrep, they were listed as "not completed". Others were
2653 picked up by some special patterns. I have improved the selection
2654 criteria to be more general.
2656 PH/38 The host_find_failed option in the manualroute router can now be set
2657 to "ignore", to completely ignore a host whose IP address cannot be
2658 found. If all hosts are ignored, the behaviour is controlled by the new
2659 host_all_ignored option.
2661 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2662 homing or because of multiple MX records with /mx) generated more than
2663 one IP address, and the following item turned out to be the local host,
2664 all the secondary addresses of the first item were incorrectly removed
2665 from the list, along with the local host and any following hosts (which
2666 is what is supposed to happen).
2668 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2669 whoever called Exim into the -H file. In the case of the daemon it was
2670 behaving confusingly. When first started, it used values for whoever
2671 started the daemon, but after a SIGHUP it used the Exim user (because it
2672 calls itself on a restart). I have changed the code so that it now always
2675 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2676 message are rejected with the same error (e.g. no authentication or bad
2677 sender address), and a DATA command is nevertheless sent (as can happen
2678 with PIPELINING or a stupid MUA), the error message that was given to the
2679 RCPT commands is included in the rejection of the DATA command. This is
2680 intended to be helpful for MUAs that show only the final error to their
2683 PH/42 Another patch from the Sieve maintainer.
2685 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2686 Eximstats - Fixed some broken HTML links and added missing column headers
2688 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2689 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2691 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2693 PH/43 Yet another patch from the Sieve maintainer.
2695 PH/44 I found a way to check for a TCP/IP connection going away before sending
2696 the response to the final '.' that terminates a message, but only in the
2697 case where the client has not sent further data following the '.'
2698 (unfortunately, this is allowed). However, in many cases there won't be
2699 any further data because there won't be any more messages to send. A call
2700 to select() can be used: if it shows that the input is "ready", there is
2701 either input waiting, or the socket has been closed. An attempt to read
2702 the next input character can distinguish the two cases. Previously, Exim
2703 would have sent an OK response which the client would never have see.
2704 This could lead to message repetition. This fix should cure that, at
2705 least in a lot of common cases.
2707 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2708 advertised in response to EHLO.
2714 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2715 fixed by 4.65/MH/01 (is this a record?) are fixed:
2717 (i) An empty string was always treated as zero by the numeric comparison
2718 operators. This behaviour has been restored.
2720 (ii) It is documented that the numeric comparison operators always treat
2721 their arguments as decimal numbers. This was broken in that numbers
2722 starting with 0 were being interpreted as octal.
2724 While fixing these problems I realized that there was another issue that
2725 hadn't been noticed. Values of message_size_limit (both the global option
2726 and the transport option) were treated as octal if they started with 0.
2727 The documentation was vague. These values are now always treated as
2728 decimal, and I will make that clear in the documentation.
2734 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2735 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2738 MH/01 Don't check that the operands of numeric comparison operators are
2739 integers when their expansion is in "skipping" mode (fixes bug
2740 introduced by 4.64-PH/07).
2742 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2743 child addresses, Exim now panics and dies. Previously, because the count
2744 is held in a short int, deliveries were likely to be lost. As such a
2745 large number of recipients for a single message is ridiculous
2746 (performance will be very, very poor), I have chosen to impose a limit
2747 rather than extend the field.
2753 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2754 leftover -K file (the existence of which was triggered by #402).
2755 While we were at it, introduced process PID as part of the -K
2756 filename. This should rule out race conditions when creating
2759 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2760 processing considerably. Previous code took too long for large mails,
2761 triggering a timeout which in turn triggers #401.
2763 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2764 in the DK code in transports.c. sendfile() is not really portable,
2765 hence the _LINUX specificness.
2767 TF/01 In the add_headers option to the mail command in an Exim filter,
2768 there was a bug that Exim would claim a syntax error in any
2769 header after the first one which had an odd number of characters
2772 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2773 callout verification, Exim cached a "reject" for the entire domain. This
2774 is correct for most verifications, but it is not correct for a recipient
2775 verification with use_sender or use_postmaster set, because in that case
2776 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2777 case of MAIL FROM:<> rejection from other early rejections (e.g.
2778 rejection of HELO). When verifying a recipient using a non-null MAIL
2779 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2780 Whatever the result of the callout, the value of the domain cache is
2781 left unchanged (for any other kind of callout, getting as far as trying
2782 RCPT means that the domain itself is ok).
2784 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2787 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2788 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2789 session, the daemon ignores ECONNECT errors and logs others; it now
2790 ignores EPIPE as well.
2792 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2793 (quoted-printable decoding).
2795 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2796 later the small subsequent patch to fix an introduced bug.
2798 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2800 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2802 PH/08 An error is now given if message_size_limit is specified negative.
2804 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2805 to be given (somewhat) arbitrary names.
2807 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2810 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2811 miscellaneous code fixes
2813 PH/10 Added the log_reject_target ACL modifier to specify where to log
2816 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2817 hostname. This is wrong, because it relates to the incoming message (and
2818 probably the interface on which it is arriving) and not to the outgoing
2819 callout (which could be using a different interface). This has been
2820 changed to use the value of the helo_data option from the smtp transport
2821 instead - this is what is used when a message is actually being sent. If
2822 there is no remote transport (possible with a router that sets up host
2823 addresses), $smtp_active_hostname is used.
2825 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2826 tweaks were necessary in order to get it to work (see also 21 below):
2827 (a) The code assumed that strncpy() returns a negative number on buffer
2828 overflow, which isn't the case. Replaced with Exim's string_format()
2830 (b) There were several signed/unsigned issues. I just did the minimum
2831 hacking in of casts. There is scope for a larger refactoring.
2832 (c) The code used strcasecmp() which is not a standard C function.
2833 Replaced with Exim's strcmpic() function.
2834 (d) The code set only $1; it now sets $auth1 as well.
2835 (e) A simple test gave the error "authentication client didn't specify
2836 service in request". It would seem that Dovecot has changed its
2837 interface. Fortunately there's a specification; I followed it and
2838 changed what the client sends and it appears to be working now.
2840 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2843 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2844 address A is aliased to B and C, where B exists and C does not. Without
2845 -v the output is "A verified" because verification stops after a
2846 successful redirection if more than one address is generated. However,
2847 with -v the child addresses are also verified. Exim was outputting "A
2848 failed to verify" and then showing the successful verification for C,
2849 with its parentage. It now outputs "B failed to verify", showing B's
2850 parentage before showing the successful verification of C.
2852 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2853 look up a TXT record in a specific list after matching in a combined
2856 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2857 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2858 they consult the DNS. I had assumed they would set it the way they
2859 wanted; and indeed my experiments on Linux seem to show that in some
2860 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2861 To be on the safe side, however, I have now made the interface to
2862 host_find_byname() similar to host_find_bydns(), with an argument
2863 containing the DNS resolver options. The host_find_byname() function now
2864 sets these options at its start, just as host_find_bydns() does. The smtp
2865 transport options dns_qualify_single and dns_search_parents are passed to
2866 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2867 of host_find_byname() use the default settings of RES_DEFNAMES
2868 (qualify_single) but not RES_DNSRCH (search_parents).
2870 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2871 spool_read_header() do less string testing, by means of a preliminary
2872 switch on the second character of optional "-foo" lines. (This is
2873 overdue, caused by the large number of possibilities that now exist.
2874 Originally there were few.) While I was there, I also converted the
2875 str(n)cmp tests so they don't re-test the leading "-" and the first
2876 character, in the hope this might squeeze out yet more improvement.
2878 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2879 flag allowing group syntax was set by the header_syntax check but not
2880 turned off, possible causing trouble later; (2) The flag was not being
2881 set at all for the header_verify test, causing "group"-style headers to
2882 be rejected. I have now set it in this case, and also caused header_
2883 verify to ignore an empty address taken from a group. While doing this, I
2884 came across some other cases where the code for allowing group syntax
2885 while scanning a header line wasn't quite right (mostly, not resetting
2886 the flag correctly in the right place). These bugs could have caused
2887 trouble for malformed header lines. I hope it is now all correct.
2889 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2890 with the "reply" argument non-NULL. The code, however (which originally
2891 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2892 but it didn't always do it. This confused somebody who was copying the
2893 code for some other use. I have removed all the tests.
2895 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2896 feature that was used to support insecure browsers during the U.S. crypto
2897 embargo. It requires special client support, and Exim is probably the
2898 only MTA that supported it -- and would never use it because real RSA is
2899 always available. This code has been removed, because it had the bad
2900 effect of slowing Exim down by computing (never used) parameters for the
2901 RSA_EXPORT functionality.
2903 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2904 authenticator to fail if there's a tab character in the incoming data
2905 (there should never be unless someone is messing about, as it's supposed
2906 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2907 if the connection is using TLS or if the remote IP is the same as the
2908 local IP, and the "valid-client-cert option" if a client certificate has
2911 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2912 authenticators. This can be used for authorization after authentication
2913 succeeds. (In the case of plaintext, it servers for both authentication
2916 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2917 if any retry times were supplied.
2919 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2920 connection, where there is no client IP address to check. In this
2921 situation, the verify now always succeeds.
2923 PH/25 Applied John Jetmore's -Mset patch.
2925 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2927 PH/27 In a string expansion for a processed (not raw) header when multiple
2928 headers of the same name were present, leading whitespace was being
2929 removed from all of them, but trailing whitespace was being removed only
2930 from the last one. Now trailing whitespace is removed from each header
2931 before concatenation. Completely empty headers in a concatenation (as
2932 before) are ignored.
2934 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2935 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2937 PH/29 [Removed. This was a change that I later backed out, and forgot to
2938 correct the ChangeLog entry (that I had efficiently created) before
2939 committing the later change.]
2941 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2942 address errors (4xx response to RCPT) over the same connection as other
2943 messages routed to the same hosts. Such deliveries are always "forced",
2944 so retry times are not inspected. This resulted in far too many retries
2945 for the affected addresses. The effect occurred only when there were more
2946 hosts than the hosts_max_try setting in the smtp transport when it had
2947 the 4xx errors. Those hosts that it had tried were not added to the list
2948 of hosts for which the message was waiting, so if all were tried, there
2949 was no problem. Two fixes have been applied:
2951 (i) If there are any address or message errors in an SMTP delivery, none
2952 of the hosts (tried or untried) are now added to the list of hosts
2953 for which the message is waiting, so the message should not be a
2954 candidate for sending over the same connection that was used for a
2955 successful delivery of some other message. This seems entirely
2956 reasonable: after all the message is NOT "waiting for some host".
2957 This is so "obvious" that I'm not sure why it wasn't done
2958 previously. Hope I haven't missed anything, but it can't do any
2959 harm, as the worst effect is to miss an optimization.
2961 (ii) If, despite (i), such a delivery is accidentally attempted, the
2962 routing retry time is respected, so at least it doesn't keep
2963 hammering the server.
2965 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2966 in ${readsocket because some servers need this prod.
2968 PH/32 Added some extra debug output when updating a wait-xxx database.
2970 PH/33 The hint "could be header name not terminated by colon", which has been
2971 given for certain expansion errors for a long time, was not being given
2972 for the ${if def:h_colon_omitted{... case.
2974 PH/34 The spec says: "With one important exception, whenever a domain list is
2975 being scanned, $domain contains the subject domain." There was at least
2976 one case where this was not true.
2978 PH/35 The error "getsockname() failed: connection reset by peer" was being
2979 written to the panic log as well as the main log, but it isn't really
2980 panic-worthy as it just means the connection died rather early on. I have
2981 removed the panic log writing for the ECONNRESET error when getsockname()
2984 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2985 runs only) independently of the message's sender address. This meant
2986 that, if the 4xx error was in fact related to the sender, a different
2987 message to the same recipient with a different sender could confuse
2988 things. In particular, this can happen when sending to a greylisting
2989 server, but other circumstances could also provoke similar problems.
2990 I have changed the default so that the retry time for these errors is now
2991 based a combination of the sender and recipient addresses. This change
2992 can be overridden by setting address_retry_include_sender=false in the
2995 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2996 remote server are returned as part of bounce messages. This was not
2997 happening for LMTP over a pipe (the lmtp transport), but now it is the
2998 same for both kinds of LMTP.
3000 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3001 in header lines that were in fact CNAMEs. This is no longer the case.
3003 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3004 and queue runs started by the daemon processed all messages. This has
3005 been fixed so that -R and -S can now usefully be given with -q<time>.
3007 PH/40 Import PCRE release 6.7 (fixes some bugs).
3009 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3011 PH/42 Give an error if -q is specified more than once.
3013 PH/43 Renamed the variables $interface_address and $interface_port as
3014 $received_ip_address and $received_port, to make it clear that these
3015 values apply to message reception, and not to the outgoing interface when
3016 a message is delivered. (The old names remain recognized, of course.)
3018 PH/44 There was no timeout on the connect() call when using a Unix domain
3019 socket in the ${readsocket expansion. There now is.
3021 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3022 be meaningful with "accept".
3024 SC/01 Eximstats V1.43
3025 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3027 SC/02 Eximstats V1.44
3028 Use a glob alias rather than an array ref in the generated
3029 parser. This improves both readability and performance.
3031 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3032 Collect SpamAssassin and rejection statistics.
3033 Don't display local sender or destination tables unless
3034 there is data to show.
3035 Added average volumes into the top table text output.
3037 SC/04 Eximstats V1.46
3038 Collect data on the number of addresses (recipients)
3039 as well as the number of messages.
3041 SC/05 Eximstats V1.47
3042 Added 'Message too big' to the list of mail rejection
3043 reasons (thanks to Marco Gaiarin).
3045 SC/06 Eximstats V1.48
3046 Mainlog lines which have GMT offsets and are too short to
3047 have a flag are now skipped.
3049 SC/07 Eximstats V1.49 (Alain Williams)
3050 Added the -emptyok flag.
3052 SC/08 Eximstats V1.50
3053 Fixes for obtaining the IP address from reject messages.
3055 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3056 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3057 whitespace changes from 4.64-PH/27
3059 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3062 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3063 are found, allow negative numbers in numeric criteria)
3065 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3067 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3068 to match changes made in 4.64-PH/43
3070 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3072 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3073 "domain" is actually an IP address, and force a failure. This locks out
3074 those revolvers/nameservers that support "A-for-A" lookups, in
3075 contravention of the specifications.
3077 PH/48 When a host name was looked up from an IP address, and the subsequent
3078 forward lookup of the name timed out, the host name was left in
3079 $sender_host_name, contrary to the specification.
3081 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3082 restricted to single-key lookups, Exim was not diagnosing an error if
3083 * or *@ was used with a query-style lookup.
3085 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3087 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3088 long ago, but noone interested enough thought of it. Let's just say that
3089 the "1.1" means that there are some new functions that weren't there at
3090 some point in the past.
3092 PH/51 Error processing for expansion failure of helo_data from an smtp
3093 transport during callout processing was broken.
3095 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3096 tested/used via the -bh/-bhc/-bs options.
3098 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3099 bug, fixed in subsequent PCRE releases).
3101 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3102 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3104 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3110 SC/01 Use a glob alias rather than an array ref in eximstats generated
3111 parser. This improves both readability and performance.
3113 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3114 Don't display local sender or destination tables in eximstats unless
3115 there is data to show.
3116 Added average volumes into the eximstats top table text output.
3118 SC/03 Collect data on the number of addresses (recipients) as well
3119 as the number of messages in eximstats.
3121 TF/01 Correct an error in the documentation for the redirect router. Exim
3122 does (usually) call initgroups() when daemonizing.
3124 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3125 with consistent privilege compared to when running as a daemon.
3127 TF/03 Note in the spec that $authenticated_id is not set for local
3128 submissions from trusted users.
3130 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3131 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3133 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3134 by adding some example configuration directives to the default
3135 configuration file. A little bit of work is required to uncomment the
3136 directives and define how usernames and passwords are checked, but
3137 there is now a framework to start from.
3139 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3140 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3141 without this. I don't know how relevant this is to other LDAP libraries.
3143 PH/02 Add the verb name to the "unknown ACL verb" error.
3145 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3147 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3149 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3150 directory not expanded when it should be if an expanded home directory
3151 was set for the address (which is overridden by the transport).
3153 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3156 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3157 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3158 because it is too late at that time, and has no effect.
3160 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3161 security issue with \' (bugzilla #107). I could not use the
3162 PQescapeStringConn() function, because it needs a PGconn value as one of
3165 PH/08 When testing addresses using -bt, indicate those final addresses that
3166 are duplicates that would not cause an additional delivery. At least one
3167 person was confused, thinking that -bt output corresponded to deliveries.
3168 (Suppressing duplicates isn't a good idea as you lose the information
3169 about possibly different redirections that led to the duplicates.)
3171 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3172 systems where poll() doesn't work, in particular OS X.
3174 PH/10 Added more information to debugging output for retry time not reached.
3176 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3177 operations in malware.c.
3179 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3182 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3183 syslog_duplication set false, log lines that would normally be written
3184 both the the main log and to the reject log were not written to syslog at
3187 PH/14 In the default configuration, change the use of "message" in ACL warn
3188 statements to "add_header".
3190 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3191 not followed by a command (e.g. "seen endif").
3193 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3194 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3197 PH/17 Added extra conditions to the default value of delay_warning_condition
3201 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3202 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3203 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3206 The Auto-Submitted: and various List- headers are standardised, whereas I
3207 don't think Precedence: ever was.
3209 PH/18 Refactored debugging code in route_finduser() to show more information,
3210 in particular, the error code if getpwnam() issues one.
3212 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3213 This is apparently needed in addition to the PH/07 change above to avoid
3214 any possible encoding problems.
3216 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3217 but not after initializing Perl.
3219 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3220 output them only if debugging. By default they are written stderr,
3221 apparently, which is not desirable.
3223 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3226 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3229 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3231 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3232 authenticated or an ident call has been made. Suppress the default
3233 values for $authenticated_id and $authenticated_sender (but permit -oMai
3234 and -oMas) when testing with -bh.
3236 PH/24 Re-jigged the order of the tests in the default configuration so that the
3237 tests for valid domains and recipients precede the DNS black list and CSA
3238 tests, on the grounds that those ones are more expensive.
3240 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3241 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3242 This bug exists in every version of Exim that I still have, right back to
3245 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3246 However, an attempt to turn on case-sensitivity in a regex key by
3247 including (?-i) didn't work because the subject string was already
3248 lowercased, and the effects were non-intuitive. It turns out that a
3249 one-line patch can be used to allow (?-i) to work as expected.
3255 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3256 other effects) broke the use of negated acl sub-conditions.
3258 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3261 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3262 "Deny" causes Exim to reject the incoming connection with a 554 error.
3263 Unfortunately, if there is a major crisis, such as a disk failure,
3264 tcp-wrappers gives "deny", whereas what one would like would be some
3265 kind of temporary error. A kludge has been added to help with this.
3266 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3267 554 error is used if errno is still zero or contains ENOENT (which occurs
3268 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3271 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3273 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3274 errors. Otherwise a message that provokes a temporary error (when other
3275 messages do not) can cause a whole host to time out.
3277 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3278 addresses were routed directly to files or pipes from a redirect router.
3279 File deliveries just didn't batch; pipe deliveries might have suffered
3282 PH/06 A failure to get a lock for a hints database would erroneously always say
3283 "Failed to get write lock", even when it was really a read lock.
3285 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3286 of 0600. This has been changed to use the value of the lockfile_mode
3287 option (which defaults to 0600).
3289 PH/08 Applied small patch from the Sieve maintainer.
3291 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3292 folder from quota calculations, a direct delivery into this folder messed
3293 up the contents of the maildirsize file. This was because the regex was
3294 used only to exclude .Trash (or whatever) when the size of the mailbox
3295 was calculated. There was no check that a delivery was happening into an
3296 excluded directory. This bug has been fixed by ignoring all quota
3297 processing for deliveries into excluded directories.
3299 PH/10 Added the maildirfolder_create_regex option to appendfile.
3305 PH/01 The code for finding all the local interface addresses on a FreeBSD
3306 system running IPv6 was broken. This may well have applied to all BSD
3307 systems, as well as to others that have similar system calls. The broken
3308 code found IPv4 interfaces correctly, but gave incorrect values for the
3309 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3310 that it would not match correctly against @[] and not recognize the IPv6
3313 PH/02 The ipliteral router was not recognizing addresses of the form user@
3314 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3316 PH/03 Added disable_ipv6.
3318 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3319 decoded form, because it is most often used to construct To: headers
3320 lines in autoreplies, and the decoded form may well be syntactically
3321 invalid. However, $reply_address has leading white space removed, and all
3322 newlines turned into spaces so that the autoreply transport does not
3325 PH/05 If group was specified without a user on a router, and no group or user
3326 was specified on a transport, the group from the router was ignored.
3328 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3329 for visible compile-time settings that can be used to change these
3330 numbers, for those that want even more. Backwards compatibility with old
3331 spool files has been maintained. However, going back to a previous Exim
3332 release will lost any variables that are in spool files.
3334 PH/07 Two small changes when running in the test harness: increase delay when
3335 passing a TCP/IP connection to a new process, in case the original
3336 process has to generate a bounce, and remove special handling of
3337 127.0.0.2 (sic), which is no longer necessary.
3339 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3340 be the same on different OS.
3342 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3345 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3346 whether --show-vars was specified or not
3348 JJ/02 exipick: Added support for new ACL variable spool format introduced
3351 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3352 syntactically invalid From: or Reply-to: line, and a filter used this to
3353 generate an autoreply, and therefore failed to obtain an address for the
3354 autoreply, Exim could try to deliver to a non-existent relative file
3355 name, causing unrelated and misleading errors. What now happens is that
3356 it logs this as a hard delivery error, but does not attempt to create a
3359 PH/11 The exinext utility has a -C option for testing purposes, but although
3360 the given file was scanned by exinext itself; it wasn't being passed on
3361 when Exim was called.
3363 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3364 an end-of-file indication when reading a command response.
3366 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3367 compiled. In many other places in Exim, IPv6 addresses are always
3368 recognized, so I have changed this. It also means that IPv4 domain
3369 literals of the form [IPV4:n.n.n.n] are now always recognized.
3371 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3372 used if the router is not running as root, for example, when verifying at
3373 ACL time, or when using -bh. The debugging output from this situation was
3374 non-existent - all you got was a failure to exec. I have made two
3377 (a) Failures to set uid/gid, the current directory, or a process leader
3378 in a subprocess such as that created by queryprogram now generate
3379 suitable debugging output when -d is set.
3381 (b) The queryprogram router detects when it is not running as root,
3382 outputs suitable debugging information if -d is set, and then runs
3383 the subprocess without attempting to change uid/gid.
3385 PH/15 Minor change to Makefile for building test_host (undocumented testing
3388 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3389 additional section of a DNS packet that returns MX or SRV records.
3390 Instead, it always explicitly searches for A/AAAA records. This avoids
3391 major problems that occur when a DNS server includes only records of one
3392 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3393 fixed another bug: if SRV records were looked up and the corresponding
3394 address records were *not* found in the additional section, the port
3395 values from the SRV records were lost.
3397 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3398 using the correct key (the original address) when searching the retry
3399 rules in order to find which one to use for generating the retry hint.
3401 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3402 adding the default one. Similarly, if it contains a Reply-To: header, the
3403 errors_reply_to option, if set, is not used.
3405 PH/19 When calculating a retry time, Exim used to measure the "time since
3406 failure" by looking at the "first failed" field in the retry record. Now
3407 it does not use this if it is later than than the arrival time of the
3408 message. Instead it uses the arrival time. This makes for better
3409 behaviour in cases where some deliveries succeed, thus re-setting the
3410 "first failed" field. An example is a quota failure for a huge message
3411 when small messages continue to be delivered. Without this change, the
3412 "time since failure" will always be short, possible causing more frequent
3413 delivery attempts for the huge message than are intended.
3414 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3416 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3417 $1, $2, $3) because the numerical variables can be reset during some
3418 expansion items (e.g. "match"), thereby losing the authentication data.
3420 PH/21 Make -bV show the size of off_t variables so that the test suite can
3421 decide whether to run tests for quotas > 2G.
3423 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3424 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3425 filecount value is greater than 2G or if a quota value is greater than 2G
3426 on a system where the size of off_t is not greater than 4, a panic error
3429 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3430 never match. The debug and -bh output now contains an explicit error
3431 message indicating a malformed IPv4 address or mask.
3433 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3434 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3435 PH/23 above applies.
3437 PH/25 Do not write to syslog when running in the test harness. The only
3438 occasion when this arises is a failure to open the main or panic logs
3439 (for which there is an explicit test).
3441 PH/26 Added the /no_tell option to "control=freeze".
3443 PH/27 If a host name lookup failed very early in a connection, for example, if
3444 the IP address matched host_lookup and the reverse lookup yielded a name
3445 that did not have a forward lookup, an error message of the form "no IP
3446 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3447 could be logged. Now it outputs the IP address instead of "NULL".
3449 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3450 allows the sender and the authenticated sender to be set when
3451 submitting a message from within Exim. Since child_open_exim() is
3452 documented for local_scan(), the new function should be too.
3454 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3455 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3456 results in an empty string is now treated as unset.
3458 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3460 PH/31 Added sender_verify_fail logging option.
3462 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3463 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3464 tidied the source and removed it altogether.
3466 PH/33 When a queue run was abandoned because the load average was too high, a
3467 log line was always written; now it is written only if the queue_run log
3468 selector is set. In addition, the log line for abandonment now contains
3469 information about the queue run such as the pid. This is always present
3470 in "start" and "stop" lines but was omitted from the "abandon" line.
3472 PH/34 Omit spaces between a header name and the colon in the error message that
3473 is given when verify = headers_syntax fails (if there are lots of them,
3474 the message gets confusing).
3476 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3477 names, as there are now some PTR records that contain slashes. This check
3478 is only to protect against broken name servers that fall over on strange
3479 characters, so the fact that it applies to all lookups doesn't matter.
3481 PH/36 Now that the new test suite is complete, we can remove some of the
3482 special code in Exim that was needed for the old test suite. For example,
3483 sorting DNS records because real resolvers return them in an arbitrary
3484 order. The new test suite's fake resolver always returns records in the
3487 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3488 bounces) except when queue_only is set, to avoid logging races between
3489 the different processes.
3491 PH/38 Panic-die if .include specifies a non-absolute path.
3493 PH/39 A tweak to the "H" retry rule from its user.
3495 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3496 a label. They prevented compilation on older perls.
3498 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3499 a warning to be raised on newish perls.
3501 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3502 on queue. Changes to match documented behaviour of showing count of
3503 messages matching specified criteria.
3505 PH/40 Changed the default ident timeout from 30s to 5s.
3507 PH/41 Added support for the use of login_cap features, on those BSD systems
3508 that have them, for controlling the resources used by pipe deliveries.
3510 PH/42 The content-scanning code uses fopen() to create files in which to put
3511 message data. Previously it was not paying any attention to the mode of
3512 the files. Exim runs with umask(0) because the rest of the code creates
3513 files with open(), and sets the required mode explicitly. Thus, these
3514 files were ending up world-writeable. This was not a big issue, because,
3515 being within the spool directory, they were not world-accessible. I have
3516 created a function called modefopen, which takes an additional mode
3517 argument. It sets umask(777), creates the file, chmods it to the required
3518 mode, then resets the umask. All the relevant calls to fopen() in the
3519 content scanning code have been changed to use this function.
3521 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3522 to 24 hours. This avoids potential overflow problems when processing G
3523 and H retry rules. I suspect nobody ever tinkers with this value.
3525 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3527 PH/45 When the plaintext authenticator is running as a client, the server's
3528 challenges are checked to ensure they are valid base64 strings. By
3529 default, the authentication attempt is cancelled if an invalid string is
3530 received. Setting client_ignore_invalid_base64 true ignores these errors.
3531 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3532 they are received. Thus, the responses can be made to depend on the
3533 challenges. If an invalid string is ignored, an empty string is placed in
3536 PH/46 Messages that are created by the autoreply transport now contains a
3537 References: header, in accordance with RFCs 2822 and 3834.
3539 PH/47 Added authenticated_sender_force to the smtp transport.
3541 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3543 PH/49 Installed latest patch from the Sieve maintainer.
3545 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3546 was also set, the appendfile transport was unnecessarily scanning a
3547 directory of message files (e.g. for maildir delivery) to find the count
3548 of files (along with the size), even though it did not need this
3549 information. It now does the scan only if it needs to find either the
3550 size of the count of files.
3552 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3554 PH/52 Two bugs concerned with error handling when the smtp transport is
3557 (i) Exim was not creating retry information for temporary errors given
3558 for individual recipients after the DATA command when the smtp transport
3559 was used in LMTP mode. This meant that they could be retried too
3560 frequently, and not timed out correctly.
3562 (ii) Exim was setting the flag that allows error details to be returned
3563 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3564 recipients that were returned after the DATA command.
3566 PH/53 This is related to PH/52, but is more general: for any failing address,
3567 when detailed error information was permitted to be returned to the
3568 sender, but the error was temporary, then after the final timeout, only
3569 "retry timeout exceeded" was returned. Now it returns the full error as
3570 well as "retry timeout exceeded".
3572 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3573 do this, and (what is worse) MTAs that accept it.
3575 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3576 will now be deprecated.
3578 PH/56 New os.c-cygwin from the Cygwin maintainer.
3580 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3581 formats (previously only available in exim formats via -bpr, -bpru,
3582 and -bpra. Now also available in native and exiqgrep formats)
3584 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3585 with very large, slow to parse queues
3587 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3589 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3591 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3592 responses to authentication challenges, though it was showing the
3593 challenges; (ii) I've removed the CR characters from the debug output for
3596 PH/58 Allow for the insertion of a newline as well as a space when a string
3597 is turned into more than one encoded-word during RFC 2047 encoding. The
3598 Sieve code now uses this.
3600 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3601 data_4xx, lost_connection, tls_required.
3603 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3604 message was being sent as an SMTP response.
3606 PH/61 Add -l and -k options to exicyclog.
3608 PH/62 When verifying, if an address was redirected to one new address, so that
3609 verification continued, and the new address failed or deferred after
3610 having set something in $address_data, the value of $address_data was not
3611 passed back to the ACL. This was different to the case when no
3612 redirection occurred. The value is now passed back in both cases.
3614 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3615 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3616 use login_cap.h, so on its own it isn't the distinguishing feature. The
3617 new name refers directly to the setclassresources() function.
3619 PH/65 Added configuration files for NetBSD3.
3621 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3623 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3624 is preferred over IPv4.
3626 PH/68 The bounce_return_message and bounce_return_body options were not being
3627 honoured for bounces generated during the reception of non-SMTP messages.
3628 In particular, this applied to messages rejected by the ACL. This bug has
3629 been fixed. However, if bounce_return_message is true and bounce_return_
3630 body is false, the headers that are returned for a non-SMTP message
3631 include only those that have been read before the error was detected.
3632 (In the case of an ACL rejection, they have all been read.)
3634 PH/69 The HTML version of the specification is now built in a directory called
3635 spec_html instead of spec.html, because the latter looks like a path with
3636 a MIME-type, and this confuses some software.
3638 PH/70 Catch two compiler warnings in sieve.c.
3640 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3641 function verify_get_ident() calls ip_connect() to connect a socket, but
3642 if the "connect()" function timed out, ip_connect() used to close the
3643 socket. However, verify_get_ident() also closes the socket later, and in
3644 between Exim writes to the log, which may get opened at this point. When
3645 the socket was closed in ip_connect(), the log could get the same file
3646 descriptor number as the socket. This naturally causes chaos. The fix is
3647 not to close the socket in ip_connect(); the socket should be closed by
3648 the function that creates it. There was only one place in the code where
3649 this was missing, in the iplookup router, which I don't think anybody now
3650 uses, but I've fixed it anyway.
3652 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3653 well as to direct DNS lookups. Otherwise the handling of names in host
3654 lists is inconsistent and therefore confusing.
3660 PH/01 Two changes to the default runtime configuration:
3662 (1) Move the checks for relay_from_hosts and authenticated clients from
3663 after to before the (commented out) DNS black list checks.
3665 (2) Add control=submission to the relay_from_hosts and authenticated
3666 clients checks, on the grounds that messages accepted by these
3667 statements are most likely to be submissions.
3669 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3671 (1) Generate an error if the third argument for the ${prvs expansion is
3674 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3677 (3) Reset the variables that are obtained from the first argument of
3678 ${prvscheck and used in the second argument before leaving the code,
3679 because their memory is reclaimed, so using them afterwards may do
3682 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3683 one (it's much easier than Tom thought :-).
3685 (5) Because of (4), we can now allow for the use of $prvscheck_result
3686 inside the third argument.
3688 PH/03 For some reason, the default setting of PATH when running a command from
3689 a pipe transport was just "/usr/bin". I have changed it to
3692 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3693 anything to be listed in the output from -bV.
3695 PH/05 When a filter generated an autoreply, the entire To: header line was
3696 quoted in the delivery log line, like this:
3698 => >A.N.Other <ano@some.domain> <original@ddress> ...
3700 This has been changed so that it extracts the operative address. There
3701 may be more than one such address. If so, they are comma-separated, like
3704 => >ano@some.domain,ona@other.domain <original@ddress> ...
3706 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3707 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3708 not being looked up in the rDNS to get a host name, Exim was showing the
3709 IP address twice in Received: lines, even though the IP addresses were
3710 identical. For example:
3712 Received: from [1.2.3.4] (helo=[1.2.3.4])
3714 However, if the real host name was known, it was omitting the HELO data
3715 if it matched the actual IP address. This has been tidied up so that it
3716 doesn't show the same IP address twice.
3718 PH/07 When both +timestamp and +memory debugging was on, the value given by
3719 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3720 called by the debug printing, thereby overwriting the timestamp buffer.
3721 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3723 PH/08 When the original message was included in an autoreply transport, it
3724 always said "this is a copy of the message, including all the headers",
3725 even if body_only or headers_only was set. It now gives an appropriate
3728 PH/09 Applied a patch from the Sieve maintainer which:
3730 o fixes some comments
3731 o adds the (disabled) notify extension core
3732 o adds some debug output for the result of if/elsif tests
3733 o points to the current vacation draft in the documentation
3734 and documents the missing references header update
3738 o fixes a bug in processing the envelope test (when testing
3739 multiple envelope elements, the last element determined the
3742 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3743 Electronic Mail") by including:
3745 Auto-submitted: auto-generated
3747 in the messages that it generates (bounce messages and others, such as
3748 warnings). In the case of bounce messages for non-SMTP messages, there was
3749 also a typo: it was using "Auto_submitted" (underscore instead of
3750 hyphen). Since every message generated by Exim is necessarily in response
3751 to another message, thes have all been changed to:
3753 Auto-Submitted: auto-replied
3755 in accordance with these statements in the RFC:
3757 The auto-replied keyword:
3759 - SHOULD be used on messages sent in direct response to another
3760 message by an automatic process,
3762 - MUST NOT be used on manually-generated messages,
3764 - MAY be used on Delivery Status Notifications (DSNs) and Message
3765 Disposition Notifications (MDNs),
3767 - MUST NOT be used on messages generated by automatic or periodic
3768 processes, except for messages which are automatic responses to
3771 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3772 to the default Received: header definition.
3774 PH/12 Added log selector acl_warn_skipped (default on).
3776 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3777 variables because (a) they are in the wrong storage pool and (b) even if
3778 they were copied, it wouldn't work properly because of the caching.
3780 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3781 checking when decoding. Apparently there are clients that generate
3782 overlong encoded strings. Why am I not surprised?
3784 PH/15 If the first argument of "${if match_address" was not empty, but did not
3785 contain an "@" character, Exim crashed. Now it writes a panic log message
3786 and treats the condition as false.
3788 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3790 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3791 extension "envelope-auth". The code is finished and in agreement with
3792 other implementations, but there is no documentation so far and in fact,
3793 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3794 not changing the active code.
3796 Print executed "if" and "elsif" statements when debugging is used. This
3797 helps a great deal to understand what a filter does.
3799 Document more things not specified clearly in RFC3028. I had all this
3800 sorted out, when out of a sudden new issues came to my mind. Oops."
3802 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3805 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3806 canonical form (as documented). However, after a host name lookup from
3807 the IP address, check_host() was doing a simple string comparison with
3808 addresses acquired from the DNS when checking that the found name did
3809 have the original IP as one of its addresses. Since any found IPv6
3810 addresses are likely to be in abbreviated form, the comparison could
3811 fail. Luckily, there already exists a function for doing the comparison
3812 by converting both addresses to binary, so now that is used instead of
3813 the text comparison.
3815 PH/20 There was another similar case to PH/19, when a complete host name was
3816 given in a host list; looking up its IP address could give an abbreviated
3817 form, whereas the current host's name might or might not be abbreviated.
3818 The same fix has been applied.
3824 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3825 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3828 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3829 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3831 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3833 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3834 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3835 Also, for "Auto-Submitted", treat anything other than "no" as
3836 non-personal, in accordance with RFC 3834. (Previously it treated
3837 anything starting "auto-" as non-personal.)
3839 TF/01 The control=submission/name=... option had a problem with syntax
3840 errors if the name included a slash character. The /name= option
3841 now slurps the rest of the string, so it can include any characters
3842 but it must come last in the list of options (after /sender_retain
3845 PH/05 Some modifications to the interface to the fake nameserver for the new
3853 TK/01 Added the "success_on_redirect" address verification option. See
3854 NewStuff for rationale and an example.
3856 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3858 PH/02 Patch to exigrep to allow it to work on syslog lines.
3860 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3861 fread() to skip over the body file's header line, because in Cygwin the
3862 header line is locked and is inaccessible.
3864 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3865 co-exist for some time) to make it clear that it is the Exim ID that is
3866 referenced, not the Message-ID: header line.
3868 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3869 string_format() function, because snprintf() does not exist on all
3872 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3873 ${stat: expansion item.
3875 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3876 protocol synchronization error", to keep the pedants happy.
3878 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3879 well as for IRIX systems, when gcc is being used. See the host.c source
3882 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3884 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3887 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3888 transport and to the smtp transport in LMTP mode.
3890 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3892 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3893 run a filter in a subprocess. This could lead to confusion in subsequent
3894 lookups in the parent process. There should also be a search_tidyup() at
3895 the end of the subprocess.
3897 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3898 only if the host matched helo_try_verify_hosts, which caused the
3899 verification to occur when the EHLO/HELO command was issued. The ACL just
3900 tested the remembered result. Now, if a previous verification attempt has
3901 not happened, "verify = helo" does it there and then.
3903 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3905 TK/03 Fix log output including CR from clamd.
3907 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3908 exist provoked a memory error which could cause a segfault.
3910 PH/15 Installed PCRE 6.2
3912 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3914 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3915 of the problem. Specifically, suggested +O2 rather than +O1 for the
3918 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3920 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3921 record, the port was not correctly logged when the outgoing_port log
3922 selector was set (it logged the transort's default port).
3924 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3925 fallback_hosts, and "hosts" in the smtp transport.
3927 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3928 host errors such as "Connection refused".
3930 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3931 authentication with radiusclient 0.4.9:
3933 - Error returned from rc_read_config was caught wrongly
3934 - Username/password not passed on to radius server due to wrong length.
3936 The presumption is that some radiusclient API changes for 4.51/PH/17
3937 were not taken care of correctly. The code is still untested by me (my
3938 Linux distribution still has 0.3.2 of radiusclient), but it was
3939 contributed by a Radius user.
3941 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3942 expanding the "port" option of the smtp transport.
3944 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3945 while reading a MIME header. Thanks to Tom Hughes for a patch.
3947 PH/24 Include config.h inside local_scan.h so that configuration settings are
3950 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3951 that in an ACL for RCPT (for example), you can examine exactly what was
3954 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3955 commands, but it was not correctly comparing the address with the actual
3956 client host address. Thus, it would show the EHLO address in Received:
3957 header lines when this was not necessary.
3959 PH/27 Added the % operator to ${eval:}.
3961 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3962 it should be ignoring failures (because with -C, for example, it has lost
3963 privilege). It wasn't ignoring creation failures other than "already
3966 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3969 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3970 because an input line was too long, either on its own, or by virtue of
3971 too many continuations, the temporary file was not being removed, and the
3972 return code was incorrect.
3974 PH/31 Missing "BOOL" in function definition in filtertest.c.
3976 PH/32 Applied Sieve patches from the maintainer.
3978 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3980 PH/33 Added "verify = not_blind".
3982 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3983 Local/Makefile (with some defaults set). These are used in built scripts
3984 such as exicyclog, but they have never been used in the exim_install
3985 script (though there are many overriding facilities there). I have
3986 arranged that the exim_install script now takes note of these two
3989 PH/35 Installed configuration files for Dragonfly.
3991 PH/36 When a locally submitted message by a trusted user did not contain a
3992 From: header, and the sender address was obtained from -f or from an SMTP
3993 MAIL command, and the trusted user did not use -F to supply a sender
3994 name, $originator_name was incorrectly used when constructing a From:
3995 header. Furthermore, $originator_name was used for submission mode
3996 messages from external hosts without From: headers in a similar way,
3997 which is clearly wrong.
3999 PH/37 Added control=suppress_local_fixups.
4001 PH/38 When log_selector = +received_sender was set, and the addition of the
4002 sender made the log line's construction buffer exactly full, or one byte
4003 less than full, an overflow happened when the terminating "\n" was
4006 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4007 when the result of a list match is failure because a DNS lookup failed.
4009 PH/40 RM_COMMAND is now used in the building process.
4011 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4012 the "build-* directories that it finds.
4014 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4015 address in a domain literal was a prefix of an interface address.
4017 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4018 when verifying a sender address, unless rewrite_headers is false.
4020 PH/44 Wrote a long comment about why errors_to addresses are verified as
4021 recipients, not senders.
4023 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4024 the ratelimit ACL was added.
4026 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4028 PH/46 Added extra information about PostgreSQL errors to the error string.
4030 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4031 avoiding the need to install special zones in a real server. This is
4032 backwards compatible; if it can't find the fake resolver, it drops back.
4033 Thus, both old and new test suites can be run.
4035 TF/02 Added util/ratelimit.pl
4037 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4038 clock is set back in time.
4040 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4041 Candler <B.Candler@pobox.com>.
4043 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4044 OK for addresses that are the result of redirections.
4046 PH/48 A number of further additions for the benefit of the new test suite,
4047 including a fake gethostbyname() that interfaces to the fake DNS resolver
4050 TF/06 The fix for widen_domains has also been applied to qualify_single and
4051 search_parents which are the other dnslookup options that can cause
4054 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4057 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4059 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4060 service" if it encountered an MX record with an empty target hostname.
4061 The message is now "an MX or SRV record indicated no SMTP service".
4063 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4064 if the DNS of the sending site is misconfigured. This is quite a
4065 common situation. This change restores the behaviour of treating a
4066 helo verification defer as a failure.
4068 PH/51 If self=fail was set on a router, the bounce message did not include the
4069 actual error message.
4075 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4077 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4078 command itself ended in error, the underlying message about the transport
4079 filter timeout was being overwritten with the pipe command error. Now the
4080 underlying error message should be appended to the second error message.
4082 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4084 PH/02 Reduce the amount of output that "make" produces by default. Full output
4085 can still be requested.
4087 PH/03 The warning log line about a condition test deferring for a "warn" verb
4088 was being output only once per connection, rather than after each
4089 occurrence (because it was using the same function as for successful
4090 "warn" verbs). This seems wrong, so I have changed it.
4092 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4093 it should not have, which might have caused a crash in the right
4094 circumstances, but probably never did.
4096 PH/04 Installed a modified version of Tony Finch's patch to make submission
4097 mode fix the return path as well as the Sender: header line, and to
4098 add a /name= option so that you can make the user's friendly name appear
4101 TF/03 Added the control = fakedefer ACL modifier.
4103 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4104 Mark Lowes for thorough testing.
4106 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4108 TK/03 Merged latest SRS patch from Miles Wilton.
4110 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4111 with the definition in sysexits.h (which is #included earlier).
4112 Fortunately, Exim does not actually use EX_OK. The code used to try to
4113 preserve the sysexits.h value, by assuming that macro definitions were
4114 scanned for macro replacements. I have been disabused of this notion,
4115 so now the code just undefines EX_OK before #including unistd.h.
4117 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4118 in the smtp transport. When a block could not be written in a single
4119 write() function, the timeout was being re-applied to each part-write.
4120 This seems wrong - if the receiver was accepting one byte at a time it
4121 would take for ever. The timeout is now adjusted when this happens. It
4122 doesn't have to be particularly precise.
4124 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4125 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4127 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4128 without a domain if the check to <postmaster@domain> fails.
4130 SC/01 Eximstats: added -xls and the ability to specify output files
4131 (patch written by Frank Heydlauf).
4133 SC/02 Eximstats: use FileHandles for outputting results.
4135 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4137 SC/04 Eximstats: fixed display of large numbers with -nvr option
4139 SC/05 Eximstats: fixed merging of reports with empty tables.
4141 SC/06 Eximstats: added the -include_original_destination flag
4143 SC/07 Eximstats: removed tabs and trailing whitespace.
4145 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4147 TK/06 MBOX spool code: Add real "From " MBOX separator line
4148 so the .eml file is really in mbox format (even though
4149 most programs do not really care). Patch from Alex Miller.
4151 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4152 The latter is generated from $received_to and is only set if the
4153 message has one envelope recipient. SA can use these headers,
4154 obviously out-of-the-box. Patch from Alex Miller.
4156 PH/08 The ${def test on a variable was returning false if the variable's
4157 value was "0", contrary to what the specification has always said!
4158 The result should be true unless the variable is empty.
4160 PH/09 The syntax error of a character other than { following "${if
4161 def:variable_name" (after optional whitespace) was not being diagnosed.
4162 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4163 accidental colon was present, for example, could give incorrect results.
4165 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4166 result is used (not including appendfile, where other changes are about
4169 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4170 This involved changing a lot of size variables from int to off_t. It
4171 should work with maildirs and everything.
4173 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4174 spamd dying while we are connected to it.
4176 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4179 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4180 function for BDB 4.3.
4182 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4184 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4185 the sources for PCRE have been re-arranged and more files are now
4188 PH/15 The code I had for printing potentially long long variables in PH/11
4189 above was not the best (it lost precision). The length of off_t variables
4190 is now inspected at build time, and an appropriate printing format (%ld
4191 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4192 to be "long long int" or "long int". This is needed for the internal
4193 formatting function string_vformat().
4195 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4196 the configuration file to be ":syslog", then the script "guesses" where
4197 the logs files are, rather than using the compiled in default. In our
4198 case the guess is not the same as the compiled default, so the script
4199 suddenly stopped working when I started to use syslog. The patch checks
4200 to see if log_file_path is "". If so, it attempts to read it from exim
4201 with no configuration file to get the compiled in version, before it
4202 falls back to the previous guessing code."
4204 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4205 implementing BATV in an Exim configuration. See NewStuff for the gory
4208 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4209 Makefile that are specific to HP-UX.
4211 PH/18 If the "use_postmaster" option was set for a recipient callout together
4212 with the "random" option, the postmaster address was used as the MAIL
4213 FROM address for the random test, but not for the subsequent recipient
4214 test. It is now used for both.
4216 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4217 patch removes a few documentation additions to RFC 3028, because the
4218 latest draft now contains them. It adds the new en;ascii-case comparator
4219 and a new error check for 8bit text in MIME parts. Comparator and
4220 require names are now matched exactly. I enabled the subaddress
4221 extension, but it is not well tested yet (read: it works for me)."
4223 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4224 rework some of the code of TK/09 above to avoid the hardwired use of
4225 "%lld" and "long long". Replaced the call to snprintf() with a call to
4228 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4229 records point to non-existent hosts", "retry timeout exceeded", and
4230 "retry time not reached for any host after a long failure period".
4232 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4233 experimental DomainKeys support:
4235 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4236 (2) On an error such as an illegally used "control", the wrong name for
4237 the control was given.
4239 These problems did NOT occur unless DomainKeys support was compiled.
4241 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4243 PH/24 Added ${if match_ip condition.
4245 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4246 databases so that it will be absolutely obvious if a crash occurs in the
4247 DB library. This is a regular occurrence (often caused by mis-matched
4250 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4251 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4252 picked up on a user's system that detects such things. There doesn't seem
4253 to be a gcc warning option for this - only an attribute that has to be
4254 put on the function's prototype. It seems that in Fedora Core 4 they have
4255 set this on a number of new functions. No doubt there will be more in due
4258 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4259 specify a transport. However, if an address that was verified by such a
4260 router was the subject of a callout, Exim crashed because it tried to
4261 read the rcpt_include_affixes from the non-existent transport. Now it
4262 just assumes that the setting of that option is false. This bug was
4263 introduced by 4.51/PH/31.
4265 PH/28 Changed -d+all to exclude +memory, because that information is very
4266 rarely of interest, but it makes the output a lot bigger. People tend to
4267 do -d+all out of habit.
4269 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4270 code in os-type was giving problems when libc.so lives in lib64, like on
4273 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4274 aren't the modern standard, and it seems that some systems' include files
4275 don't always have them. Exim was already checking for some of the newer
4276 ones like T_AAAA, and defining it itself. I've added checks for all the
4277 record types that Exim uses.
4279 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4280 not automatically generating a new one, as it is supposed to. This
4281 prevented TLS from working. If the file did exist, but contained invalid
4282 data, a new version was generated, as expected. It was only the case of a
4283 non-existent file that was broken.
4285 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4286 with a change in libdomainkeys > 0.64.
4288 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4289 from DNS. If the selector record carries the flag, it now has
4290 precedence over the domain-wide flag.
4292 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4294 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4295 the use of a port name that isn't defined in /etc/services) occurred, the
4296 message was deferred as in a normal delivery, and thus remained on the
4297 spool, instead of being failed because of the mua_wrapper setting. This
4298 is now fixed, and I tidied up some of the mua_wrapper messages at the
4301 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4302 the messages in a hash of arrays rather than using individual hashes.
4303 This is a bit cleaner and results in dramatic memory savings, albeit
4304 at a slight CPU cost.
4306 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4307 as requested by Marc Sherman.
4309 SC/10 Eximstats: added histograms for user specified patterns as requested
4312 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4314 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4315 fopen() in the content-scanning modules that did not already have it.
4321 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4322 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4324 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4326 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4328 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4329 to test Sieve filters that use "vacation".
4331 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4332 that changes the way the GnuTLS parameters are stored in the cache file.
4333 The new format can be generated externally. For backward compatibility,
4334 if the data in the cache doesn't make sense, Exim assumes it has read an
4335 old-format file, and it generates new data and writes a new file. This
4336 means that you can't go back to an older release without removing the
4339 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4340 work if there are any delivery delays because "one_time" forces the
4341 parent to be marked "delivered", so its unseen clone is never tried
4342 again. For this reason, Exim now forbids the simultaneous setting of
4345 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4346 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4347 entry if you want to know the details. The fix, however, neglected to
4348 consider the case when local delivery batching is involved. The test for
4349 "previously delivered" was not happening when checking to see if an
4350 address could be batched with a previous (undelivered) one; under
4351 certain circumstances this could lead to multiple deliveries to the same
4354 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4355 in its include files, and this causes problems building Exim.
4357 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4358 header_syntax) but Exim was just ignoring anything given after a slash.
4359 In particular, this caused confusion with an attempt to use "verify =
4360 reverse_host_lookup/defer_ok". An error is now given when options are
4361 supplied for verify items that do not have them. (Maybe reverse_host_
4362 lookup should have a defer_ok option, but that's a different point.)
4364 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4365 defined by RFC 821) to 2048, because there were problems with some AUTH
4366 commands, and RFC 1869 says the size should be increased for extended
4367 SMTP commands that take arguments.
4369 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4372 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4373 "unknown" error; now it says that the functionality isn't in the binary.
4375 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4376 an address' error message when a string expansion fails (syntax or
4377 whatever). Otherwise the password may appear in the log. Following change
4378 PH/42 below, there is no longer a chance of it appearing in a bounce
4381 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4383 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4384 of its addresses was ever tried. (Bugzilla bug #2.)
4386 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4387 the result incorrectly in the debug output. (It correctly added a newline
4388 to what was transported.)
4390 TF/01 Added $received_time.
4392 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4393 commented out examples of how to interface to a virus scanner and to
4394 SpamAssassin. Also added commented examples of av_scanner and
4395 spamd_address settings.
4397 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4398 and controls are allowed in which ACLs. There were a couple of minor
4399 errors. Some of the entries in the conditions table (which is a table of
4400 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4401 negation of where the condition IS allowed.
4403 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4405 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4406 header file does not have a version number, so I've had to invent a new
4407 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4408 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4409 radiusclient), but it was contributed by a Radius user.
4411 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4412 files or directories, for OpenSSL.
4414 PH/19 When an Exim process that is running as root has to create an Exim log
4415 file, it does so in a subprocess that runs as exim:exim so as to get the
4416 ownership right at creation (otherwise, other Exim processes might see
4417 the file with the wrong ownership). There was no test for failure of this
4418 fork() call, which would lead to the process getting stuck as it waited
4419 for a non-existent subprocess. Forks do occasionally fail when resources
4420 run out. I reviewed all the other calls to fork(); they all seem to check
4423 PH/20 When checking for unexpected SMTP input at connect time (before writing
4424 the banner), Exim was not dealing correctly with a non-positive return
4425 from the read() function. If the client had disconnected by this time,
4426 the result was a log entry for a synchronization error with an empty
4427 string after "input=" when read() returned zero. If read() returned -1
4428 (an event I could not check), uninitialized data bytes were printed.
4429 There were reports of junk text (parts of files, etc) appearing after
4432 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4434 PH/22 Added support for macro redefinition, and (re)definition in between
4435 driver and ACL definitions.
4437 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4438 forgetting to use the resulting value; it was using the unexpanded value.
4440 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4441 hadn't been configured. The fix is from Juergen Kreileder, who
4442 understands it better than I do:
4444 "Here's what I see happening with three configured cyrus_sasl
4445 authenticators configured (plain, login, cram-md5):
4447 On startup auth_cyrus_sasl_init() gets called for each of these.
4448 This means three calls to sasl_listmech() without a specified mech_list.
4449 => SASL tests which mechs of all available mechs actually work
4450 => three warnings about OTP not working
4451 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4453 With the patch, sasl_listmech() also gets called three times. But now
4454 SASL's mech_list option is set to the server_mech specified in the the
4455 authenticator. Or in other words, the answer from sasl_listmech()
4456 gets limited to just the mech you're testing for (which is different
4458 => the return list contains just 'plain' or 'login', 'cram-md5' or
4459 nothing depending on the value of ob->server_mech.
4461 I've just tested the patch: Authentication still works fine,
4462 unavailable mechs specified in the exim configuration are still
4463 caught, and the auth.log warnings about OTP are gone."
4465 PH/25 When debugging is enabled, the contents of the command line are added
4466 to the debugging output, even when log_selector=+arguments is not
4469 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4470 answer is "GNU", and only if the return is "GNU/something" is the answer
4473 PH/27 $acl_verify_message is now set immediately after the failure of a
4474 verification in an ACL, and so is available in subsequent modifiers. In
4475 particular, the message can be preserved by coding like this:
4477 warn !verify = sender
4478 set acl_m0 = $acl_verify_message
4480 Previously, $acl_verify_message was set only while expanding "message"
4481 and "log_message" when a very denied access.
4483 PH/28 Modified OS/os.c-Linux with
4485 -#ifndef OS_LOAD_AVERAGE
4486 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4488 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4489 nomenclature these days.)
4491 PH/29 Installed patch from the Sieve maintainer that adds the options
4492 sieve_useraddress and sieve_subaddress to the redirect router.
4494 PH/30 In these circumstances:
4495 . Two addresses routed to the same list of hosts;
4496 . First host does not offer TLS;
4497 . First host accepts first address;
4498 . First host gives temporary error to second address;
4499 . Second host offers TLS and a TLS session is established;
4500 . Second host accepts second address.
4501 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4502 and peerdn, if requested) that were in fact used only for the second
4505 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4506 attention to any local part prefix or suffix that was matched by the
4507 router that accepted the address. It now behaves in the same way as it
4508 does for delivery: the affixes are removed from the local part unless
4509 rcpt_include_affixes is set on the transport.
4511 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4512 timeout during the DATA phase of an incoming message.
4514 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4515 applied a patch sanctioned by the Sieve maintainer.
4517 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4518 the uid or gid is negative. A case of a negative gid caused this to be
4519 noticed. The fix allows for either to be negative.
4521 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4522 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4525 PH/36 The change PH/12 above was broken. Fixed it.
4527 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4528 the grounds that routing the same address twice would always produce the
4529 same answer. This might have been true once, but it is certainly no
4530 longer true now. Routing a child address may depend on the previous
4531 routing that produced that child. Some complicated redirection strategies
4532 went wrong when messages had multiple recipients, and made Exim's
4533 behaviour dependent on the order in which the addresses were given.
4535 I have moved the duplicate checking until after the routing is complete.
4536 Exim scans the addresses that are assigned to local and remote
4537 transports, and removes any duplicates. This means that more work will be
4538 done, as duplicates will always all be routed, but duplicates are
4539 presumably rare, so I don't expect this is of any significance.
4541 For deliveries to pipes, files, and autoreplies, the duplicate checking
4542 still happens during the routing process, since they are not going to be
4545 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4546 It corrects a timeout issue with spamd. This is Ian's comment: "The
4547 background is that sometimes spamd either never reads data from a
4548 connection it has accepted, or it never writes response data. The exiscan
4549 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4550 blindly assumes that writes won't block so it may never time out."
4552 PH/39 Allow G after quota size as well as K and M.
4554 PH/40 The value set for $authenticated_id in an authenticator may not contain
4555 binary zeroes or newlines because the value is written to log lines and
4556 to spool files. There was no check on this. Now the value is run through
4557 the string_printing() function so that such characters are converted to
4558 printable escape sequences.
4560 PH/41 $message_linecount is a new variable that contains the total number of
4561 lines in the message. Compare $body_linecount, which is the count for the
4564 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4565 bounce and delay warning messages, except in certain special cases, which
4568 (a) An SMTP error message from a remote host;
4569 (b) A message specified in a :fail: redirection;
4570 (c) A message specified in a "fail" command in a system filter;
4571 (d) A message specified in a FAIL return from the queryprogram router;
4572 (e) A message specified by the cannot_route_message router option.
4574 In these cases only, Exim does include the error details in bounce and
4575 warning messages. There are also a few cases where bland messages such
4576 as "unrouteable address" or "local delivery error" are given.
4578 PH/43 $value is now also set for the "else" part of a ${run expansion.
4580 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4581 being worked on, but at least Exim now implements the latest version to
4584 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4585 process to complete was treated as a delivery failure, a timeout while
4586 writing the message to the pipe was logged, but erroneously treated as a
4587 successful delivery. Such timeouts include transport filter timeouts. For
4588 consistency with the overall process timeout, these timeouts are now
4589 treated as errors, giving rise to delivery failures by default. However,
4590 there is now a new Boolean option for the pipe transport called
4591 timeout_defer, which, if set TRUE, converts the failures into defers for
4592 both kinds of timeout. A transport filter timeout is now identified in
4595 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4596 systems where "make" and "gmake" are different, calling "gmake" at top
4597 level broke things. I've arranged for the value of $(MAKE) to be passed
4598 from the Makefile to this script so that it can call the same version of
4602 A note about Exim versions 4.44 and 4.50
4603 ----------------------------------------
4605 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4606 changes of various kinds. As a consequence, a big documentation update was
4607 needed. This delayed the release for rather longer than seemed good, especially
4608 in the light of a couple of (minor) security issues. Therefore, the changes
4609 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4610 release. So 4.44 and 4.50 are in effect two different branches that both start
4613 I have left the 4.50 change log unchanged; it contains all the changes since
4614 4.43. The change log for 4.44 is below; many of its items are identical to
4615 those for 4.50. This seems to be the most sensible way to preserve the
4616 historical information.
4622 1. Minor wording change to the doc/README.SIEVE file.
4624 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4625 computation of the current number of files was incorrect.
4627 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4628 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4631 4. Give more explanation in the error message when the command for a transport
4632 filter fails to execute.
4634 5. There are several places where Exim runs a non-Exim command in a
4635 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4636 was being done only for the command run by the queryprogram router. It is
4637 now done for all such subprocesses. The other cases are: ${run, transport
4638 filters, and the commands run by the lmtp and pipe transports.
4640 6. Added CONFIGURE_GROUP build-time option.
4642 7. Some older OS have a limit of 256 on the maximum number of file
4643 descriptors. Exim was using setrlimit() to set 1000 as a large value
4644 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4645 systems. I've change it so that if it can't get 1000, it tries for 256.
4647 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4648 was an oversight, and furthermore, ever since the addition of extra
4649 controls (e.g. 4.43/32), the checks on when to allow different forms of
4650 "control" were broken. There should now be diagnostics for all cases when a
4651 control that does not make sense is encountered.
4653 9. Added the /retain_sender option to "control=submission".
4655 10. $recipients is now available in the predata ACL (oversight).
4657 11. Tidy the search cache before the fork to do a delivery from a message
4658 received from the command line. Otherwise the child will trigger a lookup
4659 failure and thereby defer the delivery if it tries to use (for example) a
4660 cached ldap connection that the parent has called unbind on.
4662 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4663 of $address_data from the recipient verification was clobbered by the
4664 sender verification.
4666 13. The value of address_data from a sender verification is now available in
4667 $sender_address_data in subsequent conditions in the ACL statement.
4669 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4671 15. Added a new option "connect=<time>" to callout options, to set a different
4674 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4675 was its contents. (It was OK if the option was not defined at all.)
4677 17. A "Completed" log line is now written for messages that are removed from
4678 the spool by the -Mrm option.
4680 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4681 information about exactly what failed.
4683 19. Added -dd to debug only the daemon process.
4685 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4686 handles timeouts, both on the server side and network timeouts. Renamed the
4687 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4689 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4690 It is now set to "smtps".
4692 22. $host_address is now set to the target address during the checking of
4693 ignore_target_hosts.
4695 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4696 being passed; this would have caused $sender_host_name to have been used if
4697 matching the list had actually called for a host name (not very likely,
4698 since this list is usually IP addresses). A host name is now passed as
4701 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4702 code that specifies a non-restarting handler (typically sigaction() in
4703 modern systems) in an attempt to fix a rare and obscure crash bug.
4705 25. Narrowed the window for a race in the daemon that could cause it to ignore
4706 SIGCHLD signals. This is not a major problem, because they are used only to
4707 wake it up if nothing else does.
4709 26. A malformed maildirsize file could cause Exim to calculate negative values
4710 for the mailbox size or file count. Odd effects could occur as a result.
4711 The maildirsize information is now recalculated if the size or filecount
4714 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4715 support for a long time. Removed HAVE_SYS_VFS_H.
4717 28. Installed the latest version of exipick from John Jetmore.
4719 29. In an address list, if the pattern was not a regular expression, an empty
4720 subject address (from a bounce message) matched only if the pattern was an
4721 empty string. Non-empty patterns were not even tested. This was the wrong
4722 because it is perfectly reasonable to use an empty address as part of a
4723 database query. An empty address is now tested by patterns that are
4724 lookups. However, all the other forms of pattern expect the subject to
4725 contain a local part and a domain, and therefore, for them, an empty
4726 address still always fails if the pattern is not itself empty.
4728 30. Exim went into a mad DNS loop when attempting to do a callout where the
4729 host was specified on an smtp transport, and looking it up yielded more
4730 than one IP address.
4732 31. Re-factored the code for checking spool and log partition space into a
4733 function that finds that data and another that does the check. The former
4734 is then used to implement four new variables: $spool_space, $log_space,
4735 $spool_inodes, and $log_inodes.
4737 32. The RFC2047 encoding function was originally intended for short strings
4738 such as real names; it was not keeping to the 75-character limit for
4739 encoded words that the RFC imposes. It now respects the limit, and
4740 generates multiple encoded words if necessary. To be on the safe side, I
4741 have increased the buffer size for the ${rfc2047: expansion operator from
4744 33. It is now permitted to omit both strings after an "if" condition; if the
4745 condition is true, the result is "true". As before, when the second string
4746 is omitted, a false condition yields an empty string. This makes it less
4747 cumbersome to write custom ACL and router conditions.
4749 34. Failure to deliver a bounce message always caused it to be frozen, even if
4750 there was an errors_to setting on the router. The errors_to setting is now
4753 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4754 canonical form (fully expanded) before being placed in
4755 $sender_host_address.
4757 36. The table in the code that translates DNS record types into text (T_A to
4758 "A" for instance) was missing entries for NS and CNAME. It is just possible
4759 that this could have caused confusion if both these types were looked up
4760 for the same domain, because the text type is used as part of Exim's
4761 per-process caching. But the chance of anyone hitting this buglet seems
4764 37. The dnsdb lookup has been extended in a number of ways.
4766 (1) There is a new type, "zns", which walks up the domain tree until it
4767 finds some nameserver records. It should be used with care.
4769 (2) There is a new type, "mxh", which is like "mx" except that it returns
4770 just the host names, not the priorities.
4772 (3) It is now possible to give a list of domains (or IP addresses) to be
4773 looked up. The behaviour when one of the lookups defers can be
4774 controlled by a keyword.
4776 (4) It is now possible to specify the separator character for use when
4777 multiple records are returned.
4779 38. The dnslists ACL condition has been extended: it is now possible to supply
4780 a list of IP addresses and/or domains to be looked up in a particular DNS
4783 39. Added log_selector=+queue_time_overall.
4785 40. When running the queue in the test harness, wait just a tad after forking a
4786 delivery process, to get repeatability of debugging output.
4788 41. Include certificate and key file names in error message when GnuTLS fails
4789 to set them up, because the GnuTLS error message doesn't include the name
4790 of the failing file when there is a problem reading it.
4792 42. Allow both -bf and -bF in the same test run.
4794 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4796 44. The "Exiscan patch" is now merged into the mainline Exim source.
4798 45. Sometimes the final signoff response after QUIT could fail to get
4799 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4800 < 0 before doing a fflush(). This bug looks as though it goes back to the
4801 introduction of TLS in release 3.20, but "sometimes" must have been rare
4802 because the tests only now provoked it.
4804 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4805 (this can affect the format of dates).
4807 47. exim_tidydb, when checking for the continued existence of a message for
4808 which it has found a message-specific retry record, was not finding
4809 messages that were in split spool directories. Consequently, it was
4810 deleting retry records that should have stayed in existence.
4812 48. Steve fixed some bugs in eximstats.
4814 49. The SPA authentication driver was not abandoning authentication and moving
4815 on to the next authenticator when an expansion was forced to fail,
4816 contradicting the general specification for all authenticators. Instead it
4817 was generating a temporary error. It now behaves as specified.
4819 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4820 (the order specifies the preference for clients). The order is now AES256,
4821 AES128, 3DES, ARCFOUR128.
4823 51. Small patch to Sieve code - explicitly set From: when generating an
4826 52. Exim crashed if a remote delivery caused a very long error message to be
4827 recorded - for instance if somebody sent an entire SpamAssassin report back
4828 as a large number of 550 error lines. This bug was coincidentally fixed by
4829 increasing the size of one of Exim's internal buffers (big_buffer) that
4830 happened as part of the Exiscan merge. However, to be on the safe side, I
4831 have made the code more robust (and fixed the comments that describe what
4834 53. Now that there can be additional text after "Completed" in log lines (if
4835 the queue_time_overall log selector is set), a one-byte patch to exigrep
4836 was needed to allow it to recognize "Completed" as not the last thing in
4839 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4840 patch that reportedly fixes this has been added. I am not expert enough to
4841 create a test for it. This is what the patch creator wrote:
4843 "I found a little strange behaviour of ldap code when working with
4844 Windows 2003 AD Domain, where users was placed in more than one
4845 Organization Units. When I tried to give exim partial DN, the exit code
4846 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4847 But simultaneously result of request was absolutely normal ldap result,
4848 so I produce this patch..."
4850 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4851 REFERENCE, so I have modified the code to exclude the patch when that macro
4854 55. Some experimental protocols are using DNS PTR records for new purposes. The
4855 keys for these records are domain names, not reversed IP addresses. The
4856 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4857 leaves it alone. Component reversal etc. now happens only for IP addresses.
4860 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4862 57. Double the size of the debug message buffer (to 2048) so that more of very
4863 long debug lines gets shown.
4865 58. The exicyclog utility now does better if the number of log files to keep
4866 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4868 59. Two changes related to the smtp_active_hostname option:
4870 (1) $smtp_active_hostname is now available as a variable.
4871 (2) The default for smtp_banner uses $smtp_active_hostname instead
4872 of $primary_hostname.
4874 60. The host_aton() function is supposed to be passed a string that is known
4875 to be a valid IP address. However, in the case of IPv6 addresses, it was
4876 not checking this. This is a hostage to fortune. Exim now panics and dies
4877 if the condition is not met. A case was found where this could be provoked
4878 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4879 components; fortuitously, this particular loophole had already been fixed
4880 by change 4.50/55 above.
4882 If there are any other similar loopholes, the new check in host_aton()
4883 itself should stop them being exploited. The report I received stated that
4884 data on the command line could provoke the exploit when Exim was running as
4885 exim, but did not say which command line option was involved. All I could
4886 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4887 running as the user.
4890 61. There was a buffer overflow vulnerability in the SPA authentication code
4891 (which came originally from the Samba project). I have added a test to the
4892 spa_base64_to_bits() function which I hope fixes it.
4895 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4896 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4898 63. The daemon start-up calls getloadavg() while still root for those OS that
4899 need the first call to be done as root, but it missed one case: when
4900 deliver_queue_load_max is set with deliver_drop_privilege. This is
4901 necessary for the benefit of the queue runner, because there is no re-exec
4902 when deliver_drop_privilege is set.
4904 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4905 This has been fixed.
4907 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4908 was in use, was not putting the data itself into the right store pool;
4909 consequently, it could be overwritten for a subsequent message in the same
4910 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4913 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4915 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4916 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4917 the function were treating the return as a boolean value, which happened to
4918 work because 0=false and not-0=true, but is not correct code.
4920 68. The host_aton() function was not handling scoped IPv6 addresses (those
4921 with, for example, "%eth0" on the end) correctly.
4923 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4924 negated items (that is, ~something) in unsigned ints. Some compilers
4925 apparently mutter when there is no cast.
4927 70. If an address verification called from an ACL failed, and did not produce a
4928 user-specific message (i.e. there was only a "system" message), nothing was
4929 put in $acl_verify_message. In this situation, it now puts the system
4932 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4933 session; change 4.31/43 added the unwanted input to the log line - except
4934 that it did not do this in the start of session case. It now does.
4936 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4937 This is wrong and can cause the other end to generate a synchronization
4938 error if it is another Exim or anything else that does the synchronization
4939 check. A QUIT command is no longer sent after a timeout.
4941 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4942 during host lookups.
4944 74. The defer_ok option of callout verification was not working if it was used
4945 when verifying addresses in header lines, that is, for this case:
4947 verify = header_sender/callout=defer_ok
4949 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4950 those file descriptors could be used for SMTP connections. If anything
4951 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4952 could be sent to the SMTP client, causing chaos. The daemon now opens
4953 stdin, stdout, and stderr to /dev/null when it puts itself into the
4956 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4957 log by default. The user can override this with suitable Perl magic.
4959 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4960 the log message when discard triggers, was not working for the DATA ACL or
4961 for the non-SMTP ACL.
4963 78. Error message wording change in sieve.c.
4965 79. If smtp_accept_max_per_host was set, the number of connections could be
4966 restricted to fewer than expected, because the daemon was trying to set up
4967 a new connection before checking whether the processes handling previous
4968 connections had finished. The check for completed processes is now done
4969 earlier. On busy systems, this bug wouldn't be noticed because something
4970 else would have woken the daemon, and it would have reaped the completed
4973 80. If a message was submitted locally by a user whose login name contained one
4974 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4975 It caused a spool format error. I have fixed the spool reading code. A
4976 related problem was that the "from" clause in the Received: line became
4977 illegal because of the space(s). It is now covered by ${quote_local_part.
4979 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4982 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4984 83. Patch from Sieve maintainer to fix unterminated string problem in
4985 "vacation" handling.
4987 84. Some minor changes to the Linux configuration files to help with other
4988 OS variants using glibc.
4990 85. One more patch for Sieve to update vacation handling to latest spec.
4993 ----------------------------------------------------
4994 See the note above about the 4.44 and 4.50 releases.
4995 ----------------------------------------------------
5001 1. Change 4.43/35 introduced a bug that caused file counts to be
5002 incorrectly computed when quota_filecount was set in an appendfile
5005 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5006 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5009 3. Give more explanation in the error message when the command for a transport
5010 filter fails to execute.
5012 4. There are several places where Exim runs a non-Exim command in a
5013 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5014 was being done only for the command run by the queryprogram router. It is
5015 now done for all such subprocesses. The other cases are: ${run, transport
5016 filters, and the commands run by the lmtp and pipe transports.
5018 5. Some older OS have a limit of 256 on the maximum number of file
5019 descriptors. Exim was using setrlimit() to set 1000 as a large value
5020 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5021 systems. I've change it so that if it can't get 1000, it tries for 256.
5023 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5024 was an oversight, and furthermore, ever since the addition of extra
5025 controls (e.g. 4.43/32), the checks on when to allow different forms of
5026 "control" were broken. There should now be diagnostics for all cases when a
5027 control that does not make sense is encountered.
5029 7. $recipients is now available in the predata ACL (oversight).
5031 8. Tidy the search cache before the fork to do a delivery from a message
5032 received from the command line. Otherwise the child will trigger a lookup
5033 failure and thereby defer the delivery if it tries to use (for example) a
5034 cached ldap connection that the parent has called unbind on.
5036 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5037 of $address_data from the recipient verification was clobbered by the
5038 sender verification.
5040 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5041 was its contents. (It was OK if the option was not defined at all.)
5043 11. A "Completed" log line is now written for messages that are removed from
5044 the spool by the -Mrm option.
5046 12. $host_address is now set to the target address during the checking of
5047 ignore_target_hosts.
5049 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5050 being passed; this would have caused $sender_host_name to have been used if
5051 matching the list had actually called for a host name (not very likely,
5052 since this list is usually IP addresses). A host name is now passed as
5055 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5056 code that specifies a non-restarting handler (typically sigaction() in
5057 modern systems) in an attempt to fix a rare and obscure crash bug.
5059 15. Narrowed the window for a race in the daemon that could cause it to ignore
5060 SIGCHLD signals. This is not a major problem, because they are used only to
5061 wake it up if nothing else does.
5063 16. A malformed maildirsize file could cause Exim to calculate negative values
5064 for the mailbox size or file count. Odd effects could occur as a result.
5065 The maildirsize information is now recalculated if the size or filecount
5068 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5069 support for a long time. Removed HAVE_SYS_VFS_H.
5071 18. Updated exipick to current release from John Jetmore.
5073 19. Allow an empty sender to be matched against a lookup in an address list.
5074 Previously the only cases considered were a regular expression, or an
5077 20. Exim went into a mad DNS lookup loop when doing a callout where the
5078 host was specified on the transport, if the DNS lookup yielded more than
5081 21. The RFC2047 encoding function was originally intended for short strings
5082 such as real names; it was not keeping to the 75-character limit for
5083 encoded words that the RFC imposes. It now respects the limit, and
5084 generates multiple encoded words if necessary. To be on the safe side, I
5085 have increased the buffer size for the ${rfc2047: expansion operator from
5088 22. Failure to deliver a bounce message always caused it to be frozen, even if
5089 there was an errors_to setting on the router. The errors_to setting is now
5092 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5093 canonical form (fully expanded) before being placed in
5094 $sender_host_address.
5096 24. Updated eximstats to version 1.33
5098 25. Include certificate and key file names in error message when GnuTLS fails
5099 to set them up, because the GnuTLS error message doesn't include the name
5100 of the failing file when there is a problem reading it.
5102 26. Expand error message when OpenSSL has problems setting up cert/key files.
5105 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5106 (this can affect the format of dates).
5108 28. exim_tidydb, when checking for the continued existence of a message for
5109 which it has found a message-specific retry record, was not finding
5110 messages that were in split spool directories. Consequently, it was
5111 deleting retry records that should have stayed in existence.
5113 29. eximstats updated to version 1.35
5114 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5115 1.35 - bugfix such that pie charts by volume are generated correctly
5117 30. The SPA authentication driver was not abandoning authentication and moving
5118 on to the next authenticator when an expansion was forced to fail,
5119 contradicting the general specification for all authenticators. Instead it
5120 was generating a temporary error. It now behaves as specified.
5122 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5123 (the order specifies the preference for clients). The order is now AES256,
5124 AES128, 3DES, ARCFOUR128.
5126 31. Small patch to Sieve code - explicitly set From: when generating an
5129 32. Exim crashed if a remote delivery caused a very long error message to be
5130 recorded - for instance if somebody sent an entire SpamAssassin report back
5131 as a large number of 550 error lines. This bug was coincidentally fixed by
5132 increasing the size of one of Exim's internal buffers (big_buffer) that
5133 happened as part of the Exiscan merge. However, to be on the safe side, I
5134 have made the code more robust (and fixed the comments that describe what
5137 33. Some experimental protocols are using DNS PTR records for new purposes. The
5138 keys for these records are domain names, not reversed IP addresses. The
5139 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5140 leaves it alone. Component reversal etc. now happens only for IP addresses.
5143 34. The host_aton() function is supposed to be passed a string that is known
5144 to be a valid IP address. However, in the case of IPv6 addresses, it was
5145 not checking this. This is a hostage to fortune. Exim now panics and dies
5146 if the condition is not met. A case was found where this could be provoked
5147 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5148 components; fortuitously, this particular loophole had already been fixed
5149 by change 4.50/55 or 4.44/33 above.
5151 If there are any other similar loopholes, the new check in host_aton()
5152 itself should stop them being exploited. The report I received stated that
5153 data on the command line could provoke the exploit when Exim was running as
5154 exim, but did not say which command line option was involved. All I could
5155 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5156 running as the user.
5159 35. There was a buffer overflow vulnerability in the SPA authentication code
5160 (which came originally from the Samba project). I have added a test to the
5161 spa_base64_to_bits() function which I hope fixes it.
5164 36. The daemon start-up calls getloadavg() while still root for those OS that
5165 need the first call to be done as root, but it missed one case: when
5166 deliver_queue_load_max is set with deliver_drop_privilege. This is
5167 necessary for the benefit of the queue runner, because there is no re-exec
5168 when deliver_drop_privilege is set.
5170 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5171 was in use, was not putting the data itself into the right store pool;
5172 consequently, it could be overwritten for a subsequent message in the same
5173 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5176 38. Sometimes the final signoff response after QUIT could fail to get
5177 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5178 < 0 before doing a fflush(). This bug looks as though it goes back to the
5179 introduction of TLS in release 3.20, but "sometimes" must have been rare
5180 because the tests only now provoked it.
5186 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5187 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5188 Now it returns an integer. A number of calls were still expecting a T/F
5189 return. Fortuitously, in all cases, the tests worked in OK situations,
5190 which is the norm. However, things would have gone wrong on any write
5191 failures on the smtp file descriptor. This function is used when sending
5192 messages over SMTP and also when doing verify callouts.
5194 2. When Exim is called to do synchronous delivery of a locally submitted
5195 message (the -odf or -odi options), it no longer closes stderr before doing
5198 3. Implemented the mua_wrapper option.
5200 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5202 5. Implemented the functions header_remove(), header_testname(),
5203 header_add_at_position(), and receive_remove_recipient(), and exported them
5206 6. If an ACL "warn" statement specified the addition of headers, Exim already
5207 inserted X-ACL-Warn: at the start if there was no header name. However, it
5208 was not making this test for the second and subsequent header lines if
5209 there were newlines in the string. This meant that an invalid header could
5210 be inserted if Exim was badly configured.
5212 7. Allow an ACL "warn" statement to add header lines at the start or after all
5213 the Received: headers, as well as at the end.
5215 8. Added the rcpt_4xx retry error code.
5217 9. Added postmaster_mailfrom=xxx to callout verification option.
5219 10. Added mailfrom=xxxx to the callout verification option, for verify=
5222 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5223 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5225 12. Inserted some casts to stop certain compilers warning when using pointer
5226 differences as field lengths or precisions in printf-type calls (mostly
5227 affecting debugging statements).
5229 13. Added optional readline() support for -be (dynamically loaded).
5231 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5232 same clock tick as a message's arrival, so that its received time was the
5233 same as the "first fail" time on the retry record, and that message
5234 remained on the queue past the ultimate address timeout, every queue runner
5235 would try a delivery (because it was past the ultimate address timeout) but
5236 after another failure, the ultimate address timeout, which should have then
5237 bounced the address, did not kick in. This was a "< instead of <=" error;
5238 in most cases the first failure would have been in the next clock tick
5239 after the received time, and all would be well.
5241 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5242 being recognized when the domain list was tested by the match_domain
5243 condition in an expansion string.
5245 16. Added the ${str2b64: operator.
5247 17. Exim was always calling setrlimit() to set a large limit for the number of
5248 processes, without checking whether the existing limit was already
5249 adequate. (It did check for the limit on file descriptors.) Furthermore,
5250 errors from getrlimit() and setrlimit() were being ignored. Now they are
5251 logged to the main and panic logs, but Exim does carry on, to try to do its
5252 job under whatever limits there are.
5254 18. Imported PCRE 5.0.
5256 19. Trivial typo in log message " temporarily refused connection" (the leading
5259 20. If the log selector return_path_on_delivery was set and an address was
5260 redirected to /dev/null, the delivery process crashed because it assumed
5261 that a return path would always be set for a "successful" delivery. In this
5262 case, the whole delivery is bypassed as an optimization, and therefore no
5265 21. Internal re-arrangement: the function for sending a challenge and reading
5266 a response while authentication was assuming a zero-terminated challenge
5267 string. It's now changed to take a pointer and a length, to allow for
5268 binary data in such strings.
5270 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5272 23. Exim was not respecting finduser_retries when seeking the login of the
5273 uid under which it was called; it was always trying 10 times. (The default
5274 setting of finduser_retries is zero.) Also, it was sleeping after the final
5275 failure, which is pointless.
5277 24. Implemented tls_on_connect_ports.
5279 25. Implemented acl_smtp_predata.
5281 26. If the domain in control=submission is set empty, Exim assumes that the
5282 authenticated id is a complete email address when it generates From: or
5283 Sender: header lines.
5285 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5286 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5287 chown and chgrp in /bin and hostname in /usr/bin.
5289 28. Exim was keeping the "process log" file open after each use, just as it
5290 does for the main log. This opens the possibility of it remaining open for
5291 long periods when the USR1 signal hits a daemon. Occasional processlog
5292 errors were reported, that could have been caused by this. Anyway, it seems
5293 much more sensible not to leave this file open at all, so that is what now
5296 29. The long-running daemon process does not normally write to the log once it
5297 has entered its main loop, and it closes the log before doing so. This is
5298 so that log files can straightforwardly be renamed and moved. However,
5299 there are a couple of unusual error situations where the daemon does write
5300 log entries, and I had neglected to close the log afterwards.
5302 30. The text of an SMTP error response that was received during a remote
5303 delivery was being truncated at 512 bytes. This is too short for some of
5304 the long messages that one sometimes sees. I've increased the limit to
5307 31. It is now possible to make retry rules that apply only when a message has a
5308 specific sender, in particular, an empty sender.
5310 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5311 it possible to be selective about when SMTP synchronization is enforced.
5313 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5315 32. Implemented hosts_connection_nolog.
5317 33. Added an ACL for QUIT.
5319 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5322 35. Added mailbox_size and mailbox_filecount to appendfile.
5324 36. Added control = no_multiline_responses to ACLs.
5326 37. There was a bug in the logic of the code that waits for the clock to tick
5327 in the case where the clock went backwards by a substantial amount such
5328 that the microsecond fraction of "now" was more than the microsecond
5329 fraction of "then" (but the whole seconds number was less).
5331 38. Added support for the libradius Radius client library this is found on
5332 FreeBSD (previously only the radiusclient library was supported).
5338 1. When certain lookups returned multiple values in the form name=value, the
5339 quoting of the values was not always being done properly. Specifically:
5340 (a) If the value started with a double quote, but contained no whitespace,
5342 (b) If the value contained whitespace other than a space character (i.e.
5343 tabs or newlines or carriage returns) it was not quoted.
5344 This fix has been applied to the mysql and pgsql lookups by writing a
5345 separate quoting function and calling it from the lookup code. The fix
5346 should probably also be applied to nisplus, ibase and oracle lookups, but
5347 since I cannot test any of those, I have not disturbed their existing code.
5349 2. A hit in the callout cache for a specific address caused a log line with no
5350 reason for rejecting RCPT. Now it says "Previous (cached) callout
5351 verification failure".
5353 3. There was an off-by-one bug in the queryprogram router. An over-long
5354 return line was truncated at 256 instead of 255 characters, thereby
5355 overflowing its buffer with the terminating zero. As well as fixing this, I
5356 have increased the buffer size to 1024 (and made a note to document this).
5358 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5359 when Exim is waiting for an SMTP response from a remote server, Exim
5360 restarts its select() call on the socket, thereby resetting its timeout.
5361 This is not a problem when such interrupts are rare. Somebody set up a cron
5362 job to run exiwhat every 2 minutes, which is less than the normal select()
5363 timeout (5 or 10 minutes). This meant that the select() timeout never
5364 kicked in because it was always reset. I have fixed this by comparing the
5365 time when an interrupt arrives with the time at the start of the first call
5366 to select(). If more time than the timeout has elapsed, the interrupt is
5367 treated as a timeout.
5369 5. Some internal re-factoring in preparation for the addition of Sieve
5370 extensions (by MH). In particular, the "personal" test is moved to a
5371 separate function, and given an option for scanning Cc: and Bcc: (which is
5372 not set for Exim filters).
5374 6. When Exim created an email address using the login of the caller as the
5375 local part (e.g. when creating a From: or Sender: header line), it was not
5376 quoting the local part when it contained special characters such as @.
5378 7. Installed new OpenBSD configuration files.
5380 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5381 try to make them clearer.
5383 9. Callout options, other than the timeout value, were being ignored when
5384 verifying sender addresses in header lines. For example, when using
5386 verify = header_sender/callout=no_cache
5388 the cache was (incorrectly) being used.
5390 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5391 only the Cygwin environment.
5393 11. When return_path_on_delivery was set as a log selector, if different remote
5394 addresses in the same message used different return paths and parallel
5395 remote delivery occurred, the wrong values would sometimes be logged.
5396 (Whenever a remote delivery process finished, the return path value from
5397 the most recently started remote delivery process was logged.)
5399 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5400 lines when AUTH and/or TLS are in use. This is the "received protocol"
5401 field. Exim used to use "asmtp" for authenticated SMTP, without any
5402 indication (in the protocol name) for TLS use. Now it follows the RFC and
5403 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5404 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5405 names appear in log lines as well as in Received: header lines.
5407 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5408 extensions, and comparison tests, and to fix some bugs.
5410 14. Changes to the "personal" filter test:
5412 (1) The test was buggy in that it was just doing the equivalent of
5413 "contains" tests on header lines. For example, if a user's address was
5414 anne@some.where, the "personal" test would incorrectly be true for
5416 To: susanne@some.where
5418 This test is now done by extracting each address from the header in turn,
5419 and checking the entire address. Other tests that are part of "personal"
5420 are now done using regular expressions (for example, to check local parts
5421 of addresses in From: header lines).
5423 (2) The list of non-personal local parts in From: addresses has been
5424 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5425 taken from the Sieve specification recommendations.
5427 (3) If the message contains any header line starting with "List-" it is
5428 treated as non-personal.
5430 (4) The test for "circular" in the Subject: header line has been removed
5431 because it now seems ill-conceived.
5433 15. Minor typos in src/EDITME comments corrected.
5435 16. Installed latest exipick from John Jetmore.
5437 17. If headers_add on a router specified a text string that was too long for
5438 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5439 of string_sprintf() is now avoided.
5441 18. $message_body_size was not set (it was always zero) when running the DATA
5442 ACL and the local_scan() function.
5444 19. For the "mail" command in an Exim filter, no default was being set for
5445 the once_repeat time, causing a random time value to be used if "once" was
5446 specified. (If the value happened to be <= 0, no repeat happened.) The
5447 default is now 0s, meaning "never repeat". The "vacation" command was OK
5448 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5449 (I found it when inspecting the code).
5451 20. There is now an overall timeout for performing a callout verification. It
5452 defaults to 4 times the callout timeout, which applies to individual SMTP
5453 commands during the callout. The overall timeout applies when there is more
5454 than one host that can be tried. The timeout is checked before trying the
5455 next host. This prevents very long delays if there are a large number of
5456 hosts and all are timing out (e.g. when the network connections are timing
5457 out). The value of the overall timeout can be changed by specifying an
5458 additional sub-option for "callout", called "maxwait". For example:
5460 verify = sender/callout=5s,maxwait=20s
5462 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5463 to the end before writing, but this should make it even safer).
5465 22. Exim was forgetting that it had advertised PIPELINING for the second and
5466 subsequent messages on an SMTP connection. It was also not resetting its
5467 memory on STARTTLS and an internal HELO.
5469 23. When Exim logs an SMTP synchronization error within a session, it now
5470 records whether PIPELINING has been advertised or not.
5472 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5473 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5474 rather than long int.
5476 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5478 26. Added the never_mail option to autoreply.
5484 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5485 crash if the getsockname() call failed; this can happen if a connection is
5486 closed very soon after it is established. The problem was simply in the
5487 order in which certain operations were done, causing Exim to try to write
5488 to the SMTP stream before it had set up the file descriptor. The bug has
5489 been fixed by making things happen in the correct order.
5495 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5496 before the connection was closed, thus losing the rejection response.
5498 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5499 some early Solaris releases, but causes trouble in current releases where
5500 socklen_t is defined.
5502 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5505 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5508 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5509 that releasing the top of it at the end releases what was used for sub-
5510 expansions (unless the block got too big). However, discard this block if
5511 the first thing is a variable or header, so that we can use its block when
5512 it is dynamic (useful for very large $message_headers, for example).
5514 6. Lookups now cache *every* query, not just the most recent. A new, separate
5515 store pool is used for this. It can be recovered when all lookup caches are
5516 flushed. Lookups now release memory at the end of their result strings.
5517 This has involved some general refactoring of the lookup sources.
5519 7. Some code has been added to the store_xxx() functions to reduce the amount
5520 of flapping under certain conditions.
5522 8. log_incoming_interface used to affect only the <= reception log lines. Now
5523 it causes the local interface and port to be added to several more SMTP log
5524 lines, for example "SMTP connection from", and rejection lines.
5526 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5528 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5530 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5531 could be overwritten at the end of the current message (or the start of a
5532 new message if it was set in a HELO ACL). The value is now preserved for
5533 the duration of the SMTP connection.
5535 12. If a transport had a headers_rewrite setting, and a matching header line
5536 contained an unqualified address, that address was qualified, even if it
5537 did not match any rewriting rules. The underlying bug was that the values
5538 of the flags that permit the existence of unqualified sender and recipient
5539 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5540 non-local messages, and by -bnq for local messages) were not being
5541 preserved with the message after it was received.
5543 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5544 "next input=" as part of the text comprising the host identity instead of
5545 the correct text. The code was using the same buffer for two different
5546 strings. However, depending on which order the printing function evaluated
5547 its arguments, the bug did not always show up. Under Linux, for example, my
5548 test suite worked just fine.
5550 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5551 doesn't work with some older versions of Perl. It has been changed to "my",
5552 which in any case is probably the better facility to use.
5554 15. A really picky compiler found some instances of statements for creating
5555 error messages that either had too many or two few arguments for the format
5558 16. The size of the buffer for calls to the DNS resolver has been increased
5559 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5560 for addresses that have a lot of PTR records. This alleviates a problem; it
5561 does not fully solve it.
5563 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5564 buffer now truncates the list and logs the incident, which is the same
5565 action as happens when Exim is looking up a host name and its aliases.
5566 Previously in this situation something unpredictable would happen;
5567 sometimes it was "internal error: store_reset failed".
5569 18. If a server dropped the connection unexpectedly when an Exim client was
5570 using GnuTLS and trying to read a response, the client delivery process
5571 crashed while trying to generate an error log message.
5573 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5574 string, for example:
5576 warn message = H1: something\nH2: something
5578 the text was added as a single header line from Exim's point of view
5579 though it ended up OK in the delivered message. However, searching for the
5580 second and subsequent header lines using $h_h2: did not work. This has been
5581 fixed. Similarly, if a system filter added multiple headers in this way,
5582 the routers could not see them.
5584 20. Expanded the error message when iplsearch is called with an invalid key to
5585 suggest using net-iplsearch in a host list.
5587 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5588 ACLs are no longer actually imposed (and a message to that effect is
5591 22. If a "gecos" field in a passwd entry contained escaped characters, in
5592 particular, if it contained a \" sequence, Exim got it wrong when building
5593 a From: or a Sender: header from that name. A second bug also caused
5594 incorrect handling when an unquoted " was present following a character
5595 that needed quoting.
5597 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5598 was not being matched caselessly.
5600 24. Arranged for all hyphens in the exim.8 source to be escaped with
5603 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5604 properly. Recipient callout cache records were still being keyed to include
5605 the sender, even when use_sender was set false. This led to far more
5606 callouts that were necessary. The sender is no longer included in the key
5607 when use_sender is false.
5609 26. Added "control = submission" modifier to ACLs.
5611 27. Added the ${base62d: operator to decode base 62 numbers.
5613 28. dnsdb lookups can now access SRV records.
5615 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5616 the configuration file.
5618 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5619 (-v) mode. This makes the output for a verbose queue run more intelligible.
5621 31. Added a use_postmaster feature to recipient callouts.
5623 32. Added the $body_zerocount variable, containing the number of binary zero
5624 bytes in the message body.
5626 33. The time of last modification of the "new" subdirectory is now used as the
5627 "mailbox time last read" when there is a quota error for a maildir
5630 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5632 35. Added +ignore_unknown as a special item in host lists.
5634 36. Code for decoding IPv6 addresses in host lists is now included, even if
5635 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5636 address was recognized as an IP address, but was then not correctly decoded
5637 into binary, causing unexpected and incorrect effects when compared with
5644 1. Very minor rewording of debugging text in manualroute to say "list of
5645 hosts" instead of "hostlist".
5647 2. If verify=header_syntax was set, and a header line with an unqualified
5648 address (no domain) and a large number of spaces between the end of the
5649 name and the colon was received, the reception process suffered a buffer
5650 overflow, and (when I tested it) crashed. This was caused by some obsolete
5651 code that should have been removed. The fix is to remove it!
5653 3. When running in the test harness, delay a bit after writing a bounce
5654 message to get a bit more predictability in the log output.
5656 4. Added a call to search_tidyup() just before forking a reception process. In
5657 theory, someone could use a lookup in the expansion of smtp_accept_max_
5658 per_host which, without the tidyup, could leave open a database connection.
5660 5. Added the variables $recipient_data and $sender_data which get set from a
5661 lookup success in an ACL "recipients" or "senders" condition, or a router
5662 "senders" option, similar to $domain_data and $local_part_data.
5664 6. Moved the writing of debug_print from before to after the "senders" test
5667 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5668 problems for message scanning, either using a data ACL, or using
5669 local_scan() because the Received: header was not generated till after they
5670 were called (in order to set the time as the time of reception completion).
5671 I have revised the way this works. The header is now generated after the
5672 body is received, but before the ACL or local_scan() are called. After they
5673 are run, the timestamp in the header is updated.
5679 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5680 before starting a queue runner without re-exec. This happened only when
5681 deliver_drop_privilege was set or when the Exim user was set to root. The
5682 effect of the bug was that timeouts during subsequent deliveries caused
5683 crashes instead of being properly handled. The handler is now left at its
5684 default (and expected) setting.
5686 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5687 message, again when deliver_drop_privilege is set or Exim is run as root.
5688 The bug described in (1) was not present in this case, but the tidying up
5689 of the other signals was missing. I have made the two cases consistent.
5691 3. The ignore_target_hosts setting on a manualroute router was being ignored
5692 for hosts that were looked up using the /MX notation.
5694 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5697 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5698 operated on the sender address. After changing the $sender_address to <>
5699 for the sender address verify, Exim was re-instated it as the original
5700 (before rewriting) address, but remembering that it had rewritten it, so it
5701 wasn't rewriting it again. This bug also had the effect of breaking the
5702 sender address verification caching when the sender address was rewritten.
5704 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5705 This has been changed so that if the ip literal address matches
5706 ignore_target_hosts, the router declines.
5708 7. Added expansion conditions match_domain, match_address, and match_local_
5709 part (NOT match_host).
5711 8. The placeholder for the Received: header didn't have a length field set.
5713 9. Added code to Exim itself and to exim_lock to test for a specific race
5714 condition that could lead to file corruption when using MBX delivery. The
5715 issue is with the lockfile that is created in /tmp. If this file is removed
5716 after a process has opened it but before that process has acquired a lock,
5717 there is the potential for a second process to recreate the file and also
5718 acquire a lock. This could lead to two Exim processes writing to the file
5719 at the same time. The added code performs the same test as UW imapd; it
5720 checks after acquiring the lock that its file descriptor still refers to
5721 the same named file.
5723 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5724 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5727 11. Added the smtp_active_hostname option. If used, this will typically be made
5728 to depend on the incoming interface address. Because $interface_address is
5729 not set up until the daemon has forked a reception process, error responses
5730 that can happen earlier (such as "too many connections") no longer contain
5733 12. If an expansion in a condition on a "warn" statement fails because a lookup
5734 defers, the "warn" statement is abandoned, and the next ACL statement is
5735 processed. Previously this caused the whole ACL to be aborted.
5737 13. Added the iplsearch lookup type.
5739 14. Added ident_timeout as a log selector.
5741 15. Added tls_certificate_verified as a log selector.
5743 16. Added a global option tls_require_ciphers (compare the smtp transport
5744 option of the same name). This controls incoming TLS connections.
5746 17. I finally figured out how to make tls_require_ciphers do a similar thing
5747 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5748 before starting the TLS session.
5750 18. Tabs are now shown as \t in -bP output.
5752 19. If the log selector return_path_on_delivery was set, Exim crashed when
5753 bouncing a message because it had too many Received: header lines.
5755 20. If two routers both had headers_remove settings, and the first one included
5756 a superfluous trailing colon, the final name in the first list and the
5757 first name in the second list were incorrectly joined into one item (with a
5758 colon in the middle).
5764 1. Added -C and -D options to the exinext utility, mainly to make it easier
5765 to include in the automated testing, but these could be helpful when
5766 multiple configurations are in use.
5768 2. The exinext utility was not formatting the output nicely when there was
5769 an alternate port involved in the retry record key, nor when there was a
5770 message id as well (for retries that were specific to a specific message
5771 and a specific host). It was also confused by IPv6 addresses, because of
5772 the additional colons they contain. I have fixed the IPv4 problem, and
5773 patched it up to do a reasonable job for IPv6.
5775 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5776 delivery, the log line now contains "pipelined" if PIPELINING was used.
5778 4. An SMTP transport process used to panic and die if the bind() call to set
5779 an explicit outgoing interface failed. This has been changed; it is now
5780 treated in the same way as a connect() failure.
5782 5. A reference to $sender_host_name in the part of a conditional expansion
5783 that was being skipped was still causing a DNS lookup. This no longer
5786 6. The def: expansion condition was not recognizing references to header lines
5787 that used bh_ and bheader_.
5789 7. Added the _cache feature to named lists.
5791 8. The code for checking quota_filecount in the appendfile transport was
5792 allowing one more file than it should have been.
5794 9. For compatibility with Sendmail, the command line option
5802 and sets the incoming protocol and host name (for trusted callers). The
5803 host name and its colon can be omitted when only the protocol is to be set.
5804 Note the Exim already has two private options, -pd and -ps, that refer to
5805 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5806 "s", but I don't think that's a major issue.
5808 10. A number of refactoring changes to the code, none of which should affect
5811 (a) The number of logging options was getting close to filling up the
5812 32-bit word that was used as a bit map. I have split them into two classes:
5813 those that are passed in the argument to log_write(), and those that are
5814 only ever tested independently outside of that function. These are now in
5815 separate 32-bit words, so there is plenty of room for expansion again.
5816 There is no change in the user interface or the logging behaviour.
5818 (b) When building, for example, log lines, the code previously used a
5819 macro that called string_cat() twice, in order to add two strings. This is
5820 not really sufficiently general. Furthermore, there was one instance where
5821 it was actually wrong because one of the argument was used twice, and in
5822 one call a function was used. (As it happened, calling the function twice
5823 did not affect the overall behaviour.) The macro has been replaced by a
5824 function that can join an arbitrary number of extra strings onto a growing
5827 (c) The code for expansion conditions now uses a table and a binary chop
5828 instead of a serial search (which was left over from when there were very
5829 few conditions). Also, it now recognizes conditions like "pam" even when
5830 the relevant support is not compiled in: a suitably worded error message is
5831 given if an attempt is made to use such a condition.
5833 11. Added ${time_interval:xxxxx}.
5835 12. A bug was causing one of the ddress fields not to be passed back correctly
5836 from remote delivery subprocesses. The field in question was not being
5837 subsequently used, so this caused to problems in practice.
5839 13. Added new log selectors queue_time and deliver_time.
5841 14. Might have fixed a bug in maildirsizefile handling that threw up
5842 "unexpected character" debug warnings, and recalculated the data
5843 unnecessarily. In any case, I expanded the warning message to give more
5846 15. Added the message "Restricted characters in address" to the statements in
5847 the default ACL that block characters like @ and % in local parts.
5849 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5850 Three changes have been made:
5852 (a) There was a serious bug; a negative response to MAIL caused the whole
5853 recipient domain to be cached as invalid, thereby blocking all messages
5854 to all local parts at the same domain, from all senders. This bug has
5855 been fixed. The domain is no longer cached after a negative response to
5856 MAIL if the sender used is not empty.
5858 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5861 (c) A new callout option, "use_sender" has been added for people who want
5862 the modified behaviour.
5868 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5871 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5872 indeed breaks things for older releases.
5874 3. Added additional logging to the case where there is a problem reading data
5875 from a filter that is running in a subprocess using a pipe, in order to
5876 try to track down a specific problem.
5878 4. Testing facility fudge: when running in the test harness and attempting
5879 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5880 getting "No route to host". Convert this to a timeout.
5882 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5885 6. Some OS don't have socklen_t but use size_t instead. This affects the
5886 fifth argument of getsockopt() amongst other things. This is now
5887 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5888 can be set for individual OS. I have set it for SunOS5, OSF1, and
5889 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5890 some earlier ones do not.
5892 7. Change 4.30/15 was not doing the test caselessly.
5894 8. The standard form for an IPv6 address literal was being rejected by address
5895 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5896 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5897 this, as well as the form without the "IPv6" on the front (but only when
5898 address literals are enabled, of course).
5900 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5902 10. Exim crashed if a message with an empty sender address specified by -f
5903 encountered a router with an errors_to setting. This could be provoked only
5904 by a command such as
5908 where an empty string was supplied; "<>" did not hit this bug.
5910 11. Installed PCRE release 4.5.
5912 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5913 remained set. It is now erased.
5915 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5916 times from message ids (which are base 36 rather than the normal 62).
5918 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5919 were being counted as actual protocol errors, and logged if the log
5920 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5921 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5922 and DATA following a set of rejected RCPTs do not count as protocol errors.
5923 In other words, Exim assumes they were pipelined, though this may not
5924 actually be the case. Of course, in all cases the client gets an
5925 appropriate error code.
5927 15. If a lookup fails in an ACL condition, a message about the failure may
5928 be available; it is used if testing the ACL cannot continue, because most
5929 such messages specify what the cause of the deferral is. However, some
5930 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5931 that caused an old message to be retained and used if a later statement
5932 caused a defer, replacing the real cause of the deferral.
5934 16. If an IP address had so many PTR records that the DNS lookup buffer
5935 was not large enough to hold them, Exim could crash while trying to process
5936 the truncated data. It now detects and logs this case.
5938 17. Further to 4.21/58, another change has been made: if (and only if) the
5939 first line of a message (the first header line) ends with CRLF, a bare LF
5940 in a subsequent header line has a space inserted after it, so as not to
5941 terminate the header.
5943 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5944 unnecessarily, used atoi() instead of strtol(), and didn't check the
5945 termination when getting file sizes from file names by regex.
5947 19. Completely re-implemented the support for maildirsize files, in the light
5948 of a number of problems with the previous contributed implementation
5949 (4.30/29). In particular:
5951 . If the quota is zero, the maildirsize file is maintained, but no quota is
5954 . If the maildir directory does not exist, it is created before any attempt
5955 to write a maildirsize file.
5957 . The quota value in the file is just a cache; if the quota is changed in
5958 the transport, the new value overrides.
5960 . A regular expression is available for excluding directories from the
5963 20. The autoreply transport checks the characters in options that define the
5964 message's headers; it allows continued headers, but it was checking with
5965 isspace() after an embedded newline instead of explicitly looking for a
5968 21. If all the "regular" hosts to which an address was routed had passed their
5969 expiry times, and had not reached their retry times, the address was
5970 bounced, even if fallback hosts were defined. Now Exim should go on to try
5973 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5974 equivalent code in the SMTP transport. Some hosts send humungous responses
5975 to HELO/EHLO, more than 1024 it seems.
5977 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5978 gives compiler warnings in some environments. I've now done it "properly",
5981 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5982 (because of problems with the built-in one) was declared to return uschar *
5983 instead of char *, causing compiler failure.
5985 25. Fixed a file descriptor leak when processing alias/forward files.
5987 26. Fixed a minor format string issue in dbfn.c.
5989 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5991 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5992 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5995 29. When a sender address is verified, it is cached, to save repeating the test
5996 when there is more than one recipient in a message. However, when the
5997 verification involves a callout, it is possible for different callout
5998 options to be set for different recipients. It is too complicated to keep
5999 track of this in the cache, so now Exim always runs a verification when a
6000 callout is required, relying on the callout cache for the optimization.
6001 The overhead is duplication of the address routing, but this should not be
6004 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6005 to be verified with callout=postmaster, and the main callout worked but the
6006 postmaster check failed, the verification correctly failed. However, if a
6007 subsequent RCPT command asked for sender verification *without* the
6008 postmaster check, incorrect caching caused this verification also to fail,
6011 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6012 it was not caching the DNS options (qualify_single, search_parents) that
6013 were used when the lookup failed. A subsequent lookup with different
6014 options therefore always gave the same answer, though there were cases
6015 where it should not have. (Example: a "domains = !$mx_any" option on a
6016 dnslookup router: the "domains" option is always processed without any
6017 widening, but the router might have qualify_single set.) Now Exim uses the
6018 cached value only when the same options are set.
6020 32. Added John Jetmore's "exipick" utility to the distribution.
6022 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6023 than a timeout (e.g. a certificate is required, and is not provided), an
6024 Exim server now closes the connection immediately. Previously it waited for
6025 the client to close - but if the client is SSL, it seems that they each
6026 wait for each other, leading to a delay before one of them times out.
6028 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6029 maintained 0.8.x compatibility because I don't think many are using it, and
6030 it is clearly obsolete.
6032 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6035 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6036 client certificate was expired. A simple patch fixes this, though I don't
6037 understand the full logic of why the verify callback is called multiple
6040 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6041 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6042 which causes problems with some clients (such as the Certicom SSL Plus
6043 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6044 disables the coutermeasure allowing Eudora to connect."
6046 38. Exim was not checking that a write() to a log file succeeded. This could
6047 lead to Bad Things if a log got too big, in particular if it hit a file
6048 size limit. Exim now panics and dies if it cannot write to a log file, just
6049 as it does if it cannot open a log file.
6051 39. Modified OS/Makefile-Linux so that it now contains
6053 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6055 The two -D definitions ensure that Exim is compiled with large file
6056 support, which makes it possible to handle log files that are bigger than
6059 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6060 instance) a domain was checked against a named list that involved a lookup,
6061 causing $domain_data to be set, then another domain was checked against the
6062 same list, then the first domain was re-checked, the value of $domain_data
6063 after the final check could be wrong. In particular, if the second check
6064 failed, it could be set empty. This bug probably also applied to
6067 41. The strip_trailing_dot option was not being applied to the address given
6068 with the -f command-line option.
6070 42. The code for reading a message's header from the spool was incrementing
6071 $received_count, but never initializing it. This meant that the value was
6072 incorrect (doubled) while delivering a message in the same process in which
6073 it was received. In the most common configuration of Exim, this never
6074 happens - a fresh exec is done - but it can happen when
6075 deliver_drop_privilege is set.
6077 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6078 it now includes up to 150 characters of the unexpected data in the log
6081 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6082 and building data strings. The size of both of these buffers was 10 000
6083 bytes - far larger than anybody would *ever* want, thought I. Needless to
6084 say, somebody hit the limit. I have increased the maximum line length to
6085 20 000 and the maximum data length of concatenated lines to 100 000. I have
6086 also fixed two bugs, because there was no checking on these buffers. Tsk,
6087 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6088 buffer is too small.
6090 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6091 lsearch lookups. Now it does.
6093 46. When parsing a route_list item in a manualroute router, a fixed-length
6094 buffer was used for the list of hosts. I made this 1024 bytes long,
6095 thinking that nobody would ever have a list of hosts that long. Wrong.
6096 Somebody had a whole pile of complicated expansion conditions, and the
6097 string was silently truncated, leading to an expansion error. It turns out
6098 that it is easier to change to an unlimited length (owing to other changes
6099 that have happened since this code was originally written) than to build
6100 structure for giving a limitation error. The length of the item that
6101 expands into the list of hosts is now unlimited.
6103 47. The lsearch lookup could not handle data where the length of text line was
6104 more than 4095 characters. Such lines were truncated, leading to shortened
6105 data being returned. It should now handle lines of any length.
6107 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6108 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6111 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6112 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6113 no longer happens in comment lines. A list of replacements is now placed
6114 at the head of all of the source files, except those whose only change is
6115 to replace PERL_COMMAND in the very first #! line.
6117 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6118 messages on the queue, with a bottom-up merge sort, using code contributed
6119 by Michael Haardt. This should make operations like -bp somewhat faster on
6120 large queues. It won't affect queue runners, except when queue_run_in_order
6123 51. Installed eximstats 1.31 in the distribution.
6125 52. Added support for SRV lookups to the dnslookup router.
6127 53. If an ACL referred to $message_body or $message_body_end, the value was not
6128 reset for any messages that followed in the same SMTP session.
6130 54. The store-handling optimization for building very long strings was not
6131 differentiating between the different store pools. I don't think this
6132 actually made any difference in practice, but I've tidied it.
6134 55. While running the routers to verify a sender address, $sender_address
6135 was still set to the sender address. This is wrong, because when routing to
6136 send a bounce to the sender, it would be empty. Therefore, I have changed
6137 it so that, while verifying a sender address, $sender_address is set to <>.
6138 (There is no change to what happens when verifying a recipient address.)
6140 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6141 target A or AAAA records (if not already returned) without resetting the
6142 qualify_single or search_parents options of the DNS resolver. These are
6143 inappropriate in this case because the targets of MX and SRV records must
6144 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6145 target that, when qualified, matched something in the local domain. These
6146 two options are now turned off when doing these lookups.
6148 57. It seems that at least some releases of Reiserfs (which does not have the
6149 concept of a fixed number of inodes) returns zero and not -1 for the
6150 number of available inodes. This interacted badly with check_spool_inodes,
6151 which assumed that -1 was the "no such thing" setting. What I have done is
6152 to check that the total number of inodes is greater than zero before doing
6153 the test of how many are available.
6155 58. When a "warn" ACL statement has a log_message modifier, the message is
6156 remembered, and not repeated. This is to avoid a lot of repetition when a
6157 message has many recipients that cause the same warning to be written.
6158 However, Exim was preserving the list of already written lines for an
6159 entire SMTP session, which doesn't seem right. The memory is now reset if a
6160 new message is started.
6162 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6163 debugging output unless log_rewrite was also set.
6165 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6166 of (int)(handle) when we know that handle contains (void *)(-1).
6168 61. The Exim daemon panic-logs an error return when it closes the incoming
6169 connection. However "connection reset by peer" seems to be common, and
6170 isn't really an error worthy of noting specially, so that particular error
6173 62. When Exim is trying to find all the local interfaces, it used to panic and
6174 die if the ioctl to get the interface flags failed. However, it seems that
6175 on at least one OS (Solaris 9) it is possible to have an interface that is
6176 included in the list of interfaces, but for which you get a failure error
6177 for this call. This happens when the interface is not "plumbed" into a
6178 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6179 failure of the "get flags" call assumes that the interface is down.
6181 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6182 makes life easier for people who are doing arithmetic on fields extracted
6183 from dates, where you often get leading zeros that should not be
6184 interpreted as octal.
6186 64. Added qualify_domain to the redirect router, to override the global
6189 65. If a pathologically long header line contained very many addresses (the
6190 report of this problem mentioned 10 000) and each of them was rewritten,
6191 Exim could use up a very large amount of memory. (It kept on making new
6192 copies of the header line as it rewrote, and never released the old ones.)
6193 At the expense of a bit more processing, the header rewriting function has
6194 been changed so that it no longer eats memory in this way.
6196 66. The generation of the Received: header has been moved from the time that a
6197 message starts to be received, to the time that it finishes. The timestamp
6198 in the Received: header should now be very close to that of the <= log
6199 line. There are two side-effects of this change:
6201 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6202 logged header lines no longer include the local Received: line, because
6203 it has not yet been created. The same applies to a copy of the message
6204 that is returned to a non-SMTP sender when a message is rejected.
6206 (b) When a filter file is tested using -bf, no additional Received: header
6207 is added to the test message. After some thought, I decided that this
6210 This change does not affect the value of $received_for. It is still set
6211 after address rewriting, but before local_scan() is called.
6213 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6215 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6216 gave an unhelpful panic error message, and a defer error. I have managed to
6217 change this behaviour so that it now rejects any supplied certificate,
6218 which seems right, as the list of acceptable certificates is empty.
6220 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6221 gave an unhelpful defer error. I have not managed to make this reject any
6222 supplied certificates, but the error message it gives is "no certificate
6223 supplied", which is not helpful.
6225 70. exigrep's output now also includes lines that are not associated with any
6226 message, but which match the given pattern. Implemented by a patch from
6227 Martin Sluka, which also tidied up the Perl a bit.
6229 71. Recipient callout verification, like sender verification, was using <> in
6230 the MAIL FROM command. This isn't really the right thing, since the actual
6231 sender may affect whether the remote host accepts the recipient or not. I
6232 have changed it to use the actual sender in the callout; this means that
6233 the cache record is now keyed on a recipient/sender pair, not just the
6234 recipient address. There doesn't seem to be a real danger of callout loops,
6235 since a callout by the remote host to check the sender would use <>.
6236 [SEE ABOVE: changed after hitting problems.]
6238 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6239 temporary errors. However, in the case of such a code being given after
6240 the end of a data transmission (i.e. after ".") Exim was failing to write
6241 a retry record for the message. (Yes, there was some broken host that was
6242 actually sending 8xx at this point.)
6244 73. An unknown lookup type in a host list could cause Exim to panic-die when
6245 the list was checked. (An example that provoked this was putting <; in the
6246 middle of a list instead of at the start.) If this happened during a DATA
6247 ACL check, a -D file could be left lying around. This kind of configuration
6248 error no longer causes Exim to die; instead it causes a defer error. The
6249 incident is still logged to the main and panic logs.
6251 74. Buglet left over from Exim 3 conversion. The message "too many messages
6252 in one connection" was written to the rejectlog but not the mainlog, except
6253 when address rewriting (yes!) was being logged.
6255 75. Added write_rejectlog option.
6257 76. When a system filter was run not as root (that is, when system_filter_user
6258 was set), the values of the $n variables were not being returned to the
6259 main process; thus, they were not subsequently available in the $sn
6262 77. Added +return_path_on_delivery log selector.
6264 78. A connection timeout was being treated differently from recipients deferred
6265 when testing hosts_max_try with a message that was older than the host's
6266 retry timeout. (The host should not be counted, thus allowing all hosts to
6267 be tried at least once before bouncing.) This may have been the cause of an
6268 occasionally reported bug whereby a message would remain on the queue
6269 longer than the retry timeout, but would be bounced if a delivery was
6270 forced. I say "may" because I never totally pinned down the problem;
6271 setting up timeout/retry tests is difficult. See also the next item.
6273 79. The ultimate address timeout was not being applied to errors that involved
6274 a combination of host plus message (for example, a timeout on a MAIL
6275 command). When an address resolved to a number of possible hosts, and they
6276 were not all tried for each delivery (e.g. because of hosts_max_try), a
6277 message could remain on the queue longer than the retry timeout.
6279 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6282 81. Fixed an obscure SMTP outgoing bug which required at least the following
6283 conditions: (a) there was another message waiting for the same server;
6284 (b) the server returned 5xx to all RCPT commands in the first message so
6285 that the message was not completed; (c) the server dropped the connection
6286 or gave a negative response to the RSET that Exim sends to abort the
6287 transaction. The observed case was a dropped connection after DATA that had
6288 been sent in pipelining mode. That is, the server had advertised PIPELINING
6289 but was not implementing it correctly. The effect of the bug was incorrect
6290 behaviour, such as trying another host, and this could lead to a crash.
6296 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6297 and daemon.c were passed as pointers to ints; they should have been
6298 pointers to socklen_t variables (which are typically unsigned ints).
6300 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6303 3. Fixed a really odd bug that affected only the testing scheme; patching a
6304 certain fixed string in the binary changed the value of another string that
6305 happened to be identical to the end of the original first string.
6307 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6308 name", it returns that address as the IP address. On some operating
6309 systems (e.g. Solaris), it also passes back the IP address string as the
6310 "host name". However, on others (e.g. Linux), it passes back an empty
6311 string. Exim wasn't checking for this, and was changing the host name to an
6312 empty string, assuming it had been canonicalized.
6314 5. Although rare, it is permitted to have more than one PTR record for a given
6315 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6316 all the names associated with an address, because they do in Solaris.
6317 However, it seems that they do not in Linux for data that comes from the
6318 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6319 I found this out when I moved to a new Linux workstation and tried to run
6320 the Exim test suite.
6322 To get round this problem I have changed the code so that it now does its
6323 own call to the DNS to look up PTR records when searching for a host name.
6324 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6325 addresses that are only in /etc/hosts are still found.
6327 This behaviour is, however, controlled by an option called host_lookup_
6328 order, which defaults to "bydns:byaddr". If people want to use the other
6329 order, or indeed, just use one or the other means of lookup, they can
6330 specify it in this variable.
6332 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6333 some operating systems, this comes back from gethostbyaddr() as an empty
6334 string, and this is what Exim used to test for. However, it seems that in
6335 other systems, "." is yielded. Exim now tests for this case too.
6337 7. The values of check_spool_space and check_log_space are now held internally
6338 as a number of kilobytes instead of an absolute number of bytes. If a
6339 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6340 kilobyte. This means that much larger values can be stored.
6342 8. Exim monitor: an attempt to get the action menu when not actually pointing
6343 at a message produces an empty menu entitled "No message selected". This
6344 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6345 no entries in it ("Shell widget menu has zero width and/or height"). So I
6346 have added a single, blank menu entry in this case.
6348 9. Added ${quote_local_part.
6350 10. MIME decoding is now applied to the contents of Subject: header lines when
6353 11. Now that a reference to $sender_host_address automatically causes a reverse
6354 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6355 host lookup before query-style lookups in lists that might use this
6356 variable. This has therefore been abolished, and the "net-" prefix is no
6357 longer necessary for query-style lookups.
6359 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6360 have been a typo for LFLAGS, so it has been changed.
6362 13. The install script calls Exim with "-C /dev/null" in order to find the
6363 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6364 to be output. However, since Exim outputs its version number before the
6365 error, it didn't break the script. It just looked ugly. I fixed this by
6366 always allowing "-C /dev/null" if the caller is root.
6368 14. Ignore overlarge ACL variable number when reading spool file - insurance
6369 against a later release with more variables having written the file.
6371 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6372 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6373 this, as well as the form without the "IPv6" on the front.
6375 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6376 OS/Makefile-Darwin file.
6378 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6379 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6381 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6382 variable. However, some OS use ints for inodes. I've added cast to long int
6383 to get rid of the compiler warning.
6385 19. I had forgotten to lock out "/../" in configuration file names when
6386 ALT_CONFIG_PREFIX was set.
6388 20. Routers used for verification do not need to specify transports. However,
6389 if such a router generated a host list, and callout was configured, Exim
6390 crashed, because it could not find a port number from the (non-existent)
6391 transport. It now assumes port 25 in this circumstance.
6393 21. Added the -t option to exigrep.
6395 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6396 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6397 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6398 them is set, LOOKUP_LSEARCH is forced.
6400 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6402 24. Added sender and host information to the "rejected by local_scan()" log
6403 line; previously there was no indication of these.
6405 25. Added .include_if_exists.
6407 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6408 Linux. It turns out that not all file systems support this. Apparently some
6409 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6410 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6411 not supported on the file descriptor, is now ignored when Exim is trying to
6412 sync a directory. This applies only to Linux.
6414 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6416 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6417 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6420 29. Added support for maildirsize files from supplied patch (modified a bit).
6422 30. The use of :fail: followed by an empty string could lead Exim to respond to
6423 sender verification failures with (e.g.):
6425 550 Verification failed for <xxx>
6426 550 Sender verify failed
6428 where the first response line was missing the '-' that indicates it is not
6429 the final line of the response.
6431 31. The loop for finding the name of the user that called Exim had a hardwired
6432 limit of 10; it now uses the value of finduser_retries, which is used for
6433 all other user lookups.
6435 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6438 33. Exim was neglecting to zero errno before one call of strtol() when
6439 expanding a string and expecting an integer value. On some systems this
6440 resulted in spurious "integer overflow" errors. Also, it was casting the
6441 result into an int without checking.
6443 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6444 did not work. The code looks as if it has *never* worked, though it appears
6445 to have been documented since at least release 1.62. I have made it work.
6447 35. The "timeout_DNS" error in retry rules, also documented since at least
6448 1.62, also never worked. As it isn't clear exactly what this means, and
6449 clearly it isn't a major issue, I have abolished the feature by treating it
6450 as "timeout", and writing a warning to the main and panic logs.
6452 36. The display of retry rules for -brt wasn't always showing the error code
6455 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6456 timeout_connect_A, timeout_connect_MX.
6458 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6459 to the empty sender.
6461 39. The daemon was not analysing the content of -oX till after it had closed
6462 stderr and disconnected from the controlling terminal. This meant that any
6463 syntax errors were only noted on the panic log, and the return code from
6464 the command was 0. By re-arranging the code a little, I've made the
6465 decoding happen first, so such errors now appear on stderr, and the return
6466 code is 1. However, the actual setting up of the sockets still happens in
6467 the disconnected process, so errors there are still only recorded on the
6470 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6471 connections (as happens on some IP stacks) was logged at start up time as
6472 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6473 it from "IPv6 and IPv4", which means that two separate sockets are being
6476 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6477 says whether AF_INET or AF_INET6 was passed as an argument.
6479 42. Exiwhat output was messed up when time zones were included in log
6482 43. Exiwhat now gives more information about the daemon's listening ports,
6483 and whether -tls-on-connect was used.
6485 44. The "port" option of the smtp transport is now expanded.
6487 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6488 silently ignored. Now an error message is written to the main and panic
6491 46. There's a new ACL modifier called "logwrite" which writes to a log file
6492 as soon as it is encountered.
6494 47. Added $local_user_uid and $local_user_gid at routing time.
6496 48. Exim crashed when trying to verify a sender address that was being
6499 49. Exim was recognizing only a space character after ".include". It now also
6500 recognizes a tab character.
6502 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6503 extracting the relevant information from the specification. The man page no
6504 longer contains scrambled data for the -d option, and I've added a section
6505 at the front about calling Exim under different names.
6507 51. Added "extra_headers" argument to the "mail" command in filter files.
6509 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6512 53. Installed eximstats 1.29.
6514 54. Added transport_filter_timeout as a generic transport option.
6516 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6517 Cc: header lines. This was required by RFC 822, but it not required by RFC
6520 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6521 incoming messages that did not have them. Now it does so only if the
6522 message originates locally, that is, if there is no associated remote host
6523 address. When Resent- header lines are present, this applies to the Resent-
6524 lines rather than the non-Resent- lines.
6526 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6527 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6529 58. Messages for configuration errors now include the name of the main
6530 configuration file - useful now that there may be more than one file in a
6531 list (.included file names were always shown).
6533 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6534 for those rare installations that do not start the daemon as root or run it
6535 setuid root. I've cut out the call to initgroups() if the daemon is not
6538 60. The Exim user and group can now be bound into the binary as text strings
6539 that are looked up at the start of Exim's processing.
6541 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6543 62. Added $mailstore_basename variable.
6545 63. Installed patch to sieve.c from Michael Haardt.
6547 64. When Exim failed to open the panic log after failing to open the main log,
6548 the original message it was trying to log was written to stderr and debug
6549 output, but if they were not available (the usual case in production), it
6550 was lost. Now it is written to syslog before the two lines that record the
6551 failures to open the logs.
6553 65. Users' Exim filters run in subprocesses under the user's uid. It is
6554 possible for a "deliver" command or an alias in a "personal" command to
6555 provoke an address rewrite. If logging of address rewriting is configured,
6556 this fails because the process is not running as root or exim. There may be
6557 a better way of dealing with this, but for the moment (because 4.30 needs
6558 to be released), I have disabled address rewrite logging when running a
6559 filter in a non-root, non-exim process.
6565 1. The buildconfig auxiliary program wasn't quoting the value set for
6566 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6567 not defined. This bug was masked in 4.22 by the effect that was fixed in
6570 2. Some messages that were rejected after a message id was allocated were
6571 shown as "incomplete" by exigrep. It no longer does this for messages that
6572 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6574 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6575 have allow_domain_literals set, the ID did not get logged in the <= line.
6576 Domain literals are now always recognized in Message-ID: header lines.
6578 4. The first argument for a ${extract expansion item is the key name or field
6579 number. Leading and trailing spaces in this item were not being ignored,
6580 causing some misleading effects.
6582 5. When deliver_drop_privilege was set, single queue runner processes started
6583 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6584 same command in the process it spins off) were not dropping privilege.
6586 6. When the daemon running as "exim" started a queue runner, it always
6587 re-executed Exim in the spun-off process. This is a waste of effort when
6588 deliver_drop_privilege is set. The new process now just calls the
6589 queue-runner function directly.
6595 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6598 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6599 ignored. Though the use of -oP was forcing the writing of a pid file, it
6600 was always written to the default place.
6602 3. If the message "no IP address found for host xxxx" is generated during
6603 incoming verification, it is now followed by identification of the incoming
6604 connection (so you can more easily find what provoked it).
6606 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6608 5. Added some features to "harden" Exim a bit more against certain attacks:
6610 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6611 be put in Local/Makefile. This is like the never_users runtime option,
6612 but it cannot be overridden. The default setting is "root".
6614 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6615 prefix string with which any file named in a -C command line option
6618 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6619 is retained for -C and -D only if the caller of Exim is root. Without
6620 it, the exim user may also use -C and -D and retain privilege.
6622 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6623 command line option is disabled.
6625 6. Macro names set by the -D option must start with an upper case letter, just
6626 like macro names defined in the configuration file.
6628 7. Added "dereference=" facility to LDAP.
6630 8. Two instances of the typo "uknown" in the source files are fixed.
6632 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6633 the Configure-Makefile script screwed up while processing it.
6635 10. Incorporated PCRE 4.4.
6637 11. The SMTP synchronization check was not operating right at the start of an
6638 SMTP session. For example, it could not catch a HELO sent before the client
6639 waited for the greeting. There is now a check for outstanding input at the
6640 point when the greeting is written. Because of the duplex, asynchronous
6641 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6642 way, but not yet received, when the check is performed.
6644 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6645 on TCP/IP sockets, because this apparently causes some broken clients to
6648 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6649 unchanged) from the Cygwin maintainer.
6651 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6652 was supported instead of testing for mbx. Effectively a typo.
6654 15. The spa authenticator server code was not checking that the input it
6655 received was valid base64.
6657 16. The debug output line for the "set" modifier in ACLs was not showing the
6658 name of the variable that was being set.
6660 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6662 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6663 lookup on its own. Something else was needed to trigger the lookup. For
6664 example, a match in host_lookup or the need for a host name in a host list.
6665 Now, if $sender_host_name is referenced and the host name has not yet been
6666 looked up, a lookup is performed. If the lookup fails, the variable remains
6667 empty, and $host_lookup_failed is set to "1".
6669 19. Added "eqi" as a case-independent comparison operator.
6671 20. The saslauthd authentication condition could segfault if neither service
6672 nor realm was specified.
6674 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6675 error message that was logged was misleading, and incoming SMTP
6676 connections were dropped. The message is now more accurate, and temporary
6677 errors are given to SMTP connections.
6679 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6680 (except RSET) with a 421 error, until QUIT is received. However, it was
6681 failing to send a response to QUIT.
6683 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6684 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6685 ACL is now run after the helo_try_verify_hosts code.
6687 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6688 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6689 case-independent, so other case variants are also recognized). Apparently
6690 some systems use these upper case variants.
6692 25. If more than two messages were waiting for the same host, and a transport
6693 filter was specified for the transport, Exim sent two messages over the
6694 same TCP/IP connection, and then failed with "socket operation on non-
6695 socket" when it tried to send the third.
6697 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6699 27. The extern definition of crypt16() in expand.c was not being excluded when
6700 the OS had its own crypt16() function.
6702 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6703 as a preferred synonym for return_size_limit, both as an option and as an
6706 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6708 30. Changed the default configuration ACL to relax the local part checking rule
6709 for addresses that are not in any local domains. For these addresses,
6710 slashes and pipe symbols are allowed within local parts, but the sequence
6711 /../ is explicitly forbidden.
6713 31. SPA server authentication was not clearing the challenge buffer before
6716 32. log_message in a "warn" ACL statement was writing to the reject log as
6717 well as to the main log, which contradicts the documentation and doesn't
6718 seem right (because no rejection is happening). So I have stopped it.
6720 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6721 However, I am unable to do any testing of this.
6723 34. Fixed an infelicity in the appendfile transport. When checking directories
6724 for a mailbox, to see if any needed to be created, it was accidentally
6725 using path names with one or more superfluous leading slashes; tracing
6726 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6728 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6729 contents are added to the log line that is written for every discarded
6730 recipient. (Previously a log_message setting was ignored.)
6732 36. The ${quote: operator now quotes the string if it is empty.
6734 37. The install script runs exim in order to find its version number. If for
6735 some reason other than non-existence or emptiness, which it checks, it
6736 could not run './exim', it was installing it with an empty version number,
6737 i.e. as "exim-". This error state is now caught, and the installation is
6740 38. An argument was missing from the function that creates an error message
6741 when Exim fails to connect to the socket for saslauthd authentication.
6742 This could cause Exim to crash, or give a corrupted message.
6744 39. Added isip, isip4, and isip6 to ${if conditions.
6746 40. The ACL variables $acl_xx are now saved with the message, and can be
6747 accessed later in routers, transports, and filters.
6749 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6750 strings in the file are not string-expanded.
6752 42. If a MAIL command specified a SIZE value that was too large to fit into an
6753 int variable, the check against message_size_limit failed. Such values are
6754 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6755 day this will have to be increased, but I don't think I want to be around
6756 when emails are that large.
6763 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6764 iconv() is not standard in FreeBSD.
6766 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6767 IPv6 enabled. The observed symptom was a segmentation fault on return from
6768 the function os_common_find_running_interfaces() in src/os.c.
6770 3. In the check_special_case() function in daemon.c I had used "errno" as an
6771 argument name, which causes warnings on some systems. This was basically a
6772 typo, since it was named "eno" in the comments!
6774 4. The code that waits for the clock to tick (at a resolution of some fraction
6775 of a second) so as to ensure message-id uniqueness was always waiting for
6776 at least one whole tick, when it could have waited for less. [This is
6777 almost certainly not relevant at current processor speeds, where it is
6778 unlikely to ever wait at all. But we try to future-proof.]
6780 5. The function that sleeps for a time interval that includes fractions of a
6781 second contained a race. It did not block SIGALRM between setting the
6782 timer, and suspending (a couple of lines later). If the interval was short
6783 and the sigsuspend() was delayed until after it had expired, the suspension
6784 never ended. On busy systems this could lead to processes getting stuck for
6787 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6788 process, before it forks any delivery processes. The open lookup caching
6789 mechanism meant that the open file or database connection was passed into
6790 the delivery process. The problem was that delivery processes always tidy
6791 up cached lookup data. This could cause a problem for the next delivery
6792 process started by the queue runner, because the external queue runner
6793 process does not know about the closure. So the next delivery process
6794 still has data in the lookup cache. In the case of a file lookup, there was
6795 no problem because closing a file descriptor in a subprocess doesn't affect
6796 the parent. However, if the lookup was caching a connection to a database,
6797 the connection was closed, and the second delivery process was likely to
6798 see errors such as "PGSQL: query failed: server closed the connection
6799 unexpectedly". The problem has been fixed by closing all cached lookups
6800 in a queue runner before running a delivery process.
6802 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6803 seem to have the "const" qualifier which it has on other OS. I've
6806 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6807 *of the same type* (client or server) with the same public name that an
6808 error should be diagnosed.
6810 9. When Exim looked up a host name for an IP address, but failed to find the
6811 original IP address when looking up the host name (a safety check), it
6812 output the message "<ip address> does not match any IP for NULL", which was
6813 confusing, to say the least. The bug was that the host name should have
6814 appeared instead of "NULL".
6816 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6817 user that is built into the binary, and the -C or -D options is used, root
6818 privilege is dropped before the configuration file is read. In addition,
6819 logging is switched to stderr instead of the normal log files. If the
6820 configuration then re-defines the Exim user, the unprivileged environment
6821 is probably not what is expected, so Exim logs a panic warning message (but
6824 However, if deliver_drop_privilege is set, the unprivileged state may well
6825 be exactly what is intended, so the warning has been cut out in that case,
6826 and Exim is allowed to try to write to its normal log files.
6832 1. smtp_return_error_details was not giving details for temporary sender
6833 or receiver verification errors.
6835 2. Diagnose a configuration error if two authenticators have the same public
6838 3. Exim used not to create the message log file for a message until the first
6839 delivery attempt. This could be confusing when incoming messages were held
6840 for policy or load reasons. The message log file is now created at the time
6841 the message is received, and an initial "Received" line is written to it.
6843 4. The automatically generated man page for command line options had a minor
6844 bug that caused no ill effects; however, a more serious problem was that
6845 the procedure for building the man page automatically didn't always
6846 operate. Consequently, release 4.20 contains an out-of-date version. This
6847 shouldn't happen again.
6849 5. When building Exim with embedded Perl support, the script that builds the
6850 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6851 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6853 6. The freeze_tell option was not being used for messages that were frozen on
6854 arrival, either by an ACL or by local_scan().
6856 7. Added the smtp_incomplete_transaction log selector.
6858 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6859 was accepting AUTH without a new EHLO.
6861 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6862 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6865 10. Logging of TCP/IP connections (when configured) now happens in the main
6866 daemon process instead of the child process, so that the TCP/IP connection
6867 count is more accurate (but it can never be perfect).
6869 11. The use of "drop" in a nested ACL was not being handled correctly in the
6870 outer ACL. Now, if condition failure induced by the nested "drop" causes
6871 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6872 or "require"), the connection is dropped.
6874 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6875 that yield "discard" can now be used with an "accept" or a "discard" verb,
6876 but an error is generated for any others (because I can't see a useful way
6877 to define what should happen).
6879 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6880 are now processed in the same way as lines in the Exim configuration file.
6881 In particular, continuation lines are supported.
6883 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6885 15. Added -ti meaning -t -i.
6887 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6888 domains, and warn by logging if others are found.
6890 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6891 returned from GIFCONF when Exim is trying to find the list of interfaces on
6892 a host. The code in os.c has been modified to copy each ifreq to an aligned
6893 structure in all cases.
6895 Also, in some cases, the returned ifreq's were being copied to a 'struct
6896 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6897 means the last couple of bytes of an IPv6 address could be chopped if the
6898 ifreq contained only a normal sockaddr (14 bytes storage).
6900 18. Named domain lists were not supported in the hosts_treat_as_local option.
6901 An entry such as +xxxx was not recognized, and was treated as a literal
6904 19. Ensure that header lines added by a DATA ACL are included in the reject log
6905 if the ACL subsequently rejects the message.
6907 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6908 MD5 (which is deprecated).
6910 21. When testing a filter file using -bf, Exim was writing a message when it
6911 took the sender from a "From " line in the message, but it was not doing so
6912 when it took $return_path from a Return-Path: header line. It now does.
6914 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6915 with a valid header line field name (a series of printing characters
6916 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6918 23. Changed "disc" in the source to "disk" to conform to the documentation and
6919 the book and for uniformity.
6921 24. Ignore Sendmail's -Ooption=value command line item.
6923 25. When execve() failed while trying to run a command in a pipe transport,
6924 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6925 could be confused with a return value of 69 from the command itself. This
6926 has been changed to 127, the value the shell returns if it is asked to run
6927 a non-existent command. The wording for the related log line suggests a
6928 non-existent command as the problem.
6930 26. If received_header_text expands to an empty string, do not add a Received:
6931 header line to the message. (Well, it adds a token one on the spool, but
6932 marks it "old" so that it doesn't get used or transmitted.)
6934 27. Installed eximstats 1.28 (addition of -nt option).
6936 28. There was no check for failure on the call to getsockname() in the daemon
6937 code. This can fail if there is a shortage of resources on the system, with
6938 ENOMEM, for example. A temporary error is now given on failure.
6940 29. Contrary to the C standard, it seems that in some environments, the
6941 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6942 program. Exim now does this explicitly; it affects the formatting of
6943 timestamps using strftime().
6945 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6946 complaints. I've now initialized all the variables, to avoid this.
6948 32. Header lines added by a system filter were not being "seen" during
6949 transport-time rewrites.
6951 33. The info_callback() function passed to OpenSSL is set up with type void
6952 (*)(SSL *, int, int), as described somewhere. However, when calling the
6953 function (actually a macro) that sets it up, the type void(*)() is
6954 expected. I've put in a cast to prevent warnings from picky compilers.
6956 34. If a DNS black list lookup found a CNAME record, but there were no A
6957 records associated with the domain it pointed at, Exim crashed.
6959 35. If a DNS black list lookup returned more than one A record, Exim ignored
6960 all but the first. It now scans all returned addresses if a particular IP
6961 value is being sought. In this situation, the contents of the
6962 $dnslist_value variable are a list of all the addresses, separated by a
6965 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6966 to accept a host name and all its aliases if the forward lookup for any of
6967 them yielded the IP address of the incoming connection. Now it accepts only
6968 those names whose forward lookup yields the correct IP address. Any other
6969 names are discarded. This closes a loophole whereby a rogue DNS
6970 administrator could create reverse DNS records to break through a
6971 wildcarded host restriction in an ACL.
6973 37. If a user filter or a system filter that ran in a subprocess used any of
6974 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6975 the wrong values were passed to the pipe command ($thisaddress had the
6976 value of $0, $0 had the value of $1, etc). This bug was introduced by
6977 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6979 38. Improved the line breaking for long SMTP error messages from ACLs.
6980 Previously, if there was no break point between 40 and 75 characters, Exim
6981 left the rest of the message alone. Two changes have been made: (a) I've
6982 reduced the minimum length to 35 characters; (b) if it can't find a break
6983 point between 35 and 75 characters, it looks ahead and uses the first one
6984 that it finds. This may give the occasional overlong line, but at least the
6985 remaining text gets split now.
6987 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6988 file descriptors might be low, and that setting 1000 would always raise it.
6989 It turns out that in some environments, the limit is already over 1000 and
6990 that lowering it causes trouble. So now Exim takes care not to decrease it.
6992 40. When delivering a message, the value of $return_path is set to $sender_
6993 address at the start of routing (routers may change the value). By an
6994 oversight, this default was not being set up when an address was tested by
6995 -bt or -bv, which affected the outcome if any router or filter referred to
6998 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6999 log only when "message" or "log_message" are set. However, if one of the
7000 conditions was an address verification, or a call to a nested ACL, the
7001 messages generated by the underlying test were being passed through. This
7002 no longer happens. The underlying message is available in $acl_verify_
7003 message for both "message" and "log_message" expansions, so it can be
7004 passed through if needed.
7006 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7007 new expansion $bh_ to give the encoded byte string without charset
7008 translation. Translation happens only if iconv() is available; HAVE_ICONV
7009 indicates this at build time. HEADERS_CHARSET gives the charset to
7010 translate to; headers_charset can change it in the configuration, and
7011 "headers charset" can change it in an individual filter file.
7013 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7014 that creates RFC 2047 encoded "words" labels them as that charset instead
7015 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7016 expansion operator; (ii) when Exim creates a From: line for a local
7017 message; (iii) when a header line is rewritten to include a "phrase" part.
7019 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7020 buggy, causing it to skip the first lines of messages whose message ID
7021 ended in 'D'. This would not have bitten before Exim release 4.14, because
7022 message IDs were unlikely to end in 'D' before then. The effect was to have
7023 incorrect size information for certain domains.
7025 45. #include "config.h" was missing at the start of the crypt16.c module. This
7026 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7029 46. If there was a timeout during a "random" callout check, Exim treated it as
7030 a failure of the random address, and carried on sending RSET and the real
7031 address. If the delay was just some slowness somewhere, the response to the
7032 original RCPT would be taken as a response to RSET and so on, causing
7033 mayhem of various kinds.
7035 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7036 when I implemented it. It didn't allow for the fact that some option values
7037 may legitimately be negative (e.g. size_addition), and it didn't even do
7038 the right test for positive values.
7040 48. Domain names in DNS records are case-independent. Exim always looks them up
7041 in lower case. Some resolvers return domain names in exactly the case they
7042 appear in the zone file, that is, they may contain uppercase letters. Not
7043 all resolvers do this - some return always lower case. Exim was treating a
7044 change of case by a resolver as a change of domain, similar to a widening
7045 of a domain abbreviation. This triggered its re-routing code and so it was
7046 trying to route what was effectively the same domain again. This normally
7047 caused routing to fail (because the router wouldn't handle the domain
7048 twice). Now Exim checks for this case specially, and just changes the
7049 casing of the domain that it ultimately uses when it transmits the message
7052 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7055 50. If a filter generated a file delivery with a non-absolute name (possible if
7056 no home directory exists for the router), the forbid_file option was not
7059 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7060 the existing equality matching.
7062 52. Exim was using ints instead of ino_t variables in some places where it was
7063 dealing with inode numbers.
7065 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7066 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7067 called TMPDIR, and if it finds it is different, it changes its value.
7069 54. The smtp_printf() function is now made available to local_scan() so
7070 additional output lines can be written before returning. There is also an
7071 smtp_fflush() function to enable the detection of a dropped connection.
7072 The variables smtp_input and smtp_batched_input are exported to
7075 55. Changed the default runtime configuration: the message "Unknown user"
7076 has been removed from the ACL, and instead placed on the localuser router,
7077 using the cannot_route_message feature. This means that any verification
7078 failures that generate their own messages won't get overridden. Similarly,
7079 the "Unrouteable address" message that was in the ACL for unverifiable
7080 relay addresses has also been removed.
7082 56. Added hosts_avoid_esmtp to the smtp transport.
7084 57. The exicyclog script was not checking for the esoteric option
7085 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7086 will work only if exicyclog is run under the appropriate euid.
7088 58. Following a discussion on the list, the rules by which Exim recognises line
7089 endings on incoming messages have been changed. The -dropcr and drop_cr
7090 options are now no-ops, retained only for backwards compatibility. The
7091 following line terminators are recognized: LF CRLF CR. However, special
7092 processing applies to CR:
7094 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7095 nor a local message in the state where . is a terminator.
7097 (ii) If a bare CR is encountered in a header line, an extra space is added
7098 after the line terminator so as not to end the header. The reasoning
7099 behind this is that bare CRs in header lines are most likely either
7100 to be mistakes, or people trying to play silly games.
7102 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7103 was being incorrectly given as 18 bytes larger than it should have been.
7104 This is a VOB (very old bug).
7106 60. This may never have affected anything current, but just in case it has:
7107 When the local host is found other than at the start of a list of hosts,
7108 the local host, those with the same MX, and any that follow, are discarded.
7109 When the list in question was part of a longer list of hosts, the following
7110 hosts (not currently being processed) were also being discarded. This no
7111 longer happens. I'm not sure if this situation could ever has previously
7114 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7117 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7118 header. This is something that is recommended in a new Internet Draft, and
7119 is something that is documented as being done by Sendmail. There are two
7120 possible values. For messages generated by the autoreply transport, Exim
7123 Auto-Submitted: auto-replied
7125 whereas for all other generated messages (e.g. bounces) it adds
7127 Auto-Submitted: auto-generated
7129 63. The "personal" condition in filters now includes a test for the
7130 Auto-Submitted: header. If it contains the string "auto-" the message it
7131 not considered personal.
7133 64. Added rcpt_include_affixes as a generic transport option.
7135 65. Added queue_only_override (default true).
7137 66. Added the syslog_duplication option.
7139 67. If what should have been the first header line of a message consisted of
7140 a space followed by a colon, Exim was mis-interpreting it as a header line.
7141 It isn't of course - it is syntactically invalid and should therefore be
7142 treated as the start of the message body. The misbehaviour could have
7143 caused a number of strange effects, including loss of data in subsequent
7144 header lines, and spool format errors.
7146 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7147 client host had authenticated. This control can now be exercised by an ACL
7148 for more flexibility.
7150 69. By default, callouts do not happen when testing with -bh. There is now a
7151 variant, -bhc, which does actually run the callout code, including
7152 consulting and updating the callout cache.
7154 70. Added support for saslauthd authentication, courtesy of Alexander
7157 71. If statvfs() failed on the spool or log directories while checking their
7158 size for availability, Exim confusingly gave the error "space shortage".
7159 Furthermore, in debugging mode it crashed with a floating point exception.
7160 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7161 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7162 serious problem, Exim now writes to the main and panic logs when this
7163 happens, with details of the failure. It then refuses to accept the
7164 incoming message, giving the message "spool directory problem" or "log
7165 directory problem" with a 421 code for SMTP messages.
7167 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7168 0, 1, and 2 exist, because some OS complain for execs without them (see
7169 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7170 descriptors. However, the code omitted to check that the open succeeded,
7171 causing mysterious errors if for some reason the permissions on /dev/null
7172 got screwed. Now Exim writes a message to the main and panic logs, and
7173 bombs out if it can't open /dev/null.
7175 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7176 interact so that it is all more flexible. It is supposed to remain
7177 backwards compatible. Also added extra_local_interfaces.
7179 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7180 to bomb out with an assertion failure - to the client this appears as a
7181 connection drop. This problem occurs in the part of the code that was taken
7182 from the Samba project. Fortunately, the assertion is in a very simple
7183 function, so I have fixed this by reproducing the function inline in the
7184 one place where it is called, and arranging for authentication to fail
7185 instead of killing the process with assert().
7187 75. The SPA client code was not working when the server requested OEM rather
7188 than Unicode encoding.
7190 76. Added code to make require_files with a specific uid setting more usable in
7191 the case where statting the file as root fails - usually a non-root-mounted
7192 NFS file system. When this happens and the failure is EACCES, Exim now
7193 forks a subprocess and does the per-uid checking as the relevant uid.
7195 77. Added process_log_path.
7197 78. If log_file_path was not explicitly set, a setting of check_log_space or
7198 check_log_inodes was ignored.
7200 79. If a space check for the spool or log partitions fails, the incident is now
7201 logged. Of course, in the latter case the data may get lost...
7203 80. Added the %p formatting code to string_format() so that it can be used to
7204 print addresses in debug_print(). Adjusted all the address printing in the
7205 debugging in store.c to use %p rather than %d.
7207 81. There was a concern that a line of code in smtp_in.c could overflow a
7208 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7209 initially expressed, the concern was not well-founded, because trailing
7210 spaces are removed early. However, if the trailing spaces were followed by
7211 a NULL, they did not get removed, so the overflow was possible. Two fixes
7214 (a) I re-wrote the offending code in a cleaner fashion.
7215 (b) If an incoming SMTP command contains a NULL character, it is rejected
7218 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7219 runs initgroups(), so that if the Exim user is in any additional groups,
7220 they will be used during message reception.
7226 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob