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.
57 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
58 When redis_servers is set to a list of > 1 element, and the Redis servers
59 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
60 case of MOVED into a DEFER case instead, thus moving the query onto the
61 next server in the list. For a cluster of N elements, all N servers must
62 be defined in redis_servers.
64 GF/02 Catch and remove uninitialized value warning in exiqsumm
65 Check for existence of @ARGV before looking at $ARGV[0]
67 JH/01 Replace the store_release() internal interface with store_newblock(),
68 which internalises the check required to safely use the old one, plus
69 the allocate and data copy operations duplicated in both (!) of the
72 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
73 modifier. This matches the restriction on the commandline.
75 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
76 Previously only the last row was returned.
78 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
79 we assumed that tags in the header were well-formed, and parsed the
80 element content after inspecting only the first char of the tag.
81 Assumptions at that stage could crash the receive process on malformed
84 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
85 While running the DKIM ACL we operate on the Permanent memory pool so that
86 variables created with "set" persist to the DATA ACL. Also (at any time)
87 DNS lookups that fail create cache records using the Permanent pool. But
88 expansions release any allocations made on the current pool - so a dnsdb
89 lookup expansion done in the DKIM ACL releases the memory used for the
90 DNS negative-cache, and bad things result. Solution is to switch to the
91 Main pool for expansions.
92 While we're in that code, add checks on the DNS cache during store_reset,
93 active in the testsuite.
94 Problem spotted, and debugging aided, by Wolfgang Breyha.
96 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
97 When none of the hosts presented to a transport match an already-open
98 connection, close it and proceed with the list. Previously we would
99 queue the message. Spotted by Lena with Yahoo, probably involving
102 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
103 Previously a spurious "250 OK id=" response was appended to the proper
106 JH/08 The "support for" informational output now, which built with Content
107 Scanning support, has a line for the malware scanner interfaces compiled
108 in. Interface can be individually included or not at build time.
110 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
111 by the template makefile "src/EDITME". The "STREAM" support for an older
112 ClamAV interface method is removed.
114 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
115 rows affected is given instead).
117 JH/11 The runtime Berkeley DB library version is now additionally output by
118 "exim -d -bV". Previously only the compile-time version was shown.
120 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
121 SMTP connection. Previously, when one had more receipients than the
122 first, an abortive onward connection was made. Move to full support for
123 multiple onward connections in sequence, handling cutthrough connection
124 for all multi-message initiating connections.
126 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
127 routers. Previously, a multi-recipient message would fail to match the
128 onward-connection opened for the first recipient, and cause its closure.
130 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
131 a timeout on read on a GnuTLS initiating connection, resulting in the
132 initiating connection being dropped. This mattered most when the callout
133 was marked defer_ok. Fix to keep the two timeout-detection methods
136 JH/15 Relax results from ACL control request to enable cutthrough, in
137 unsupported situations, from error to silently (except under debug)
138 ignoring. This covers use with PRDR, frozen messages, queue-only and
141 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
143 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
144 metadata, resulting in a crash in free().
146 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
147 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
148 Broken also in d185889f4, with init system revamp.
150 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
151 Previously we abruptly closed the connection after reading a malware-
152 found indication; now we go on to read the "scan ok" response line,
155 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
156 ACL. Previously, a crash would result.
158 JH/19 Speed up macro lookups during configuration file read, by skipping non-
159 macro text after a replacement (previously it was only once per line) and
160 by skipping builtin macros when searching for an uppercase lead character.
162 JH/20 DANE support moved from Experimental to mainline. The Makefile control
163 for the build is renamed.
165 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
166 was allocated for every new TLS startup, meaning one per message. Fix
167 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
169 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
170 reported the original. Fix to report (as far as possible) the ACL
171 result replacing the original.
173 JH/23 Fix memory leak during multi-message connections using STARTTLS under
174 OpenSSL. Certificate information is loaded for every new TLS startup,
175 and the resources needed to be freed.
177 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
179 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
182 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
183 DATA response info to the (existing) per-recipient response info for
184 the "C=" log element. It can have useful tracking info from the
185 destination system. Patch from Simon Arlott.
187 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
188 length value. Previously this would segfault.
190 HS/02 Support Avast multiline protoocol, this allows passing flags to
191 newer versions of the scanner.
193 JH/28 Ensure that variables possibly set during message acceptance are marked
194 dead before release of memory in the daemon loop. This stops complaints
195 about them when the debug_store option is enabled. Discovered specifically
196 for sender_rate_period, but applies to a whole set of variables.
197 Do the same for the queue-runner and queue-list loops, for variables set
198 from spool message files. Do the same for the SMTP per-message loop, for
199 certain variables indirectly set in ACL operations.
201 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
202 as a multi-recipient message from a mailinglist manager). The coding had
203 an arbitrary cutoff number of characters while checking for more input;
204 enforced by writing a NUL into the buffer. This corrupted long / fast
205 input. The problem was exposed more widely when more pipelineing of SMTP
206 responses was introduced, and one Exim system was feeding another.
207 The symptom is log complaints of SMTP syntax error (NUL chars) on the
208 receiving system, and refused recipients seen by the sending system
209 (propating to people being dropped from mailing lists).
210 Discovered and pinpointed by David Carter.
212 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
213 replaced by the ${authresults } expansion.
215 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
217 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
218 allows proper process termination in container environments.
220 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
221 Previously the "final dot" had a newline after it; ensure it is CR,LF.
223 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
224 and "err_perm", deprecated since 4.83 when the RFC-defined words
225 "temperror" and "permerror" were introduced.
227 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
228 transport-filters or DKIM-signing. The restriction was lost in the
229 consolidation of verify-callout and delivery SMTP handling.
230 Extend the restriction to also cover ARC-signing.
232 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
233 in defer=pass mode supply a 450 to the initiator. Previously the message
236 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
237 tls_require_ciphers is used as before.
239 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
240 "pass_unscanned". Only tmpfails from the scanner are written to
241 the paniclog, as they may require admin intervention (permission
242 denied, license issues). Other scanner errors (like decompression
243 bombs) do not cause a paniclog entry.
245 JH/36 Fix reinitialisation of DKIM logging variable between messages.
246 Previously it was possible to log spurious information in receive log
249 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
250 triggered odd behaviour from Outlook Express clients.
252 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
255 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
256 since the IETF WG has not yet settled on that versus the original
257 "bare" representation.
259 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
260 Previously the millisecond value corrupted the output.
261 Fix also for syslog_pid=no and log_selector +pid, for which the pid
262 corrupted the output.
268 JH/01 Rework error string handling in TLS interface so that the caller in
269 more cases is responsible for logging. This permits library-sourced
270 string to be attached to addresses during delivery, and collapses
271 pairs of long lines into single ones.
273 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
274 during configuration. Wildcards are allowed and expanded.
276 JH/02 Rework error string handling in DKIM to pass more info back to callers.
277 This permits better logging.
279 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
280 do not close it down and have the child transport start it up again on
281 the passed-on TCP connection. Instead, proxy the child (and any
282 subsequent ones) for TLS via a unix-domain socket channel. Logging is
283 affected: the continued delivery log lines do not have any DNSSEC, TLS
284 Certificate or OCSP information. TLS cipher information is still logged.
286 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
287 identical IP addresses on different listening ports. Will also affect
290 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
291 add noisy ifdef guards to special-case this sillyness.
292 Patch from Bernd Kuhls.
294 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
295 than 255 are no longer allowed.
297 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
298 Disable the session-cache too, which might reduce our load. Since we
299 currrectly use a new context for every connection, both as server and
300 client, there is no benefit for these.
301 GnuTLS appears to not support tickets server-side by default (we don't
302 call gnutls_session_ticket_enable_server()) but client side is enabled
303 by default on recent versions (3.1.3 +) unless the PFS priority string
306 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
307 <https://reproducible-builds.org/specs/source-date-epoch/>.
309 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
310 the check for any unsuccessful recipients did not notice the limit, and
311 erroneously found still-pending ones.
313 JH/08 Pipeline CHUNKING command and data together, on kernels that support
314 MSG_MORE. Only in-clear (not on TLS connections).
316 JH/09 Avoid using a temporary file during transport using dkim. Unless a
317 transport-filter is involved we can buffer the headers in memory for
318 creating the signature, and read the spool data file once for the
319 signature and again for transmission.
321 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
322 4.77 as the kernel support then wasn't solid, having issues in 64bit
323 mode. Now, it's been long enough. Add support for FreeBSD also.
325 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
326 case where the routing stage had gathered several addresses to send to
327 a host before calling the transport for the first, we previously failed
328 to close down TLS in the old transport process before passing the TCP
329 connection to the new process. The new one sent a STARTTLS command
330 which naturally failed, giving a failed delivery and bloating the retry
331 database. Investigation and fix prototype from Wolfgang Breyha.
333 JH/12 Fix check on SMTP command input synchronisation. Previously there were
334 false-negatives in the check that the sender had not preempted a response
335 or prompt from Exim (running as a server), due to that code's lack of
336 awareness of the SMTP input buffering.
338 PP/04 Add commandline_checks_require_admin option.
339 Exim drops privileges sanely, various checks such as -be aren't a
340 security problem, as long as you trust local users with access to their
341 own account. When invoked by services which pass untrusted data to
342 Exim, this might be an issue. Set this option in main configuration
343 AND make fixes to the calling application, such as using `--` to stop
346 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
347 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
348 responses to those, into a single TLS record each way (this usually means
349 a single packet). As a side issue, smtp_enforce_sync now works on TLS
352 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
353 affects you only if you're dancing at the edge of the param size limits.
354 If you are, and this message makes sense to you, then: raise the
355 configured limit or use OpenSSL 1.1. Nothing we can do for older
358 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
359 cmdline element to get the documented default one. Previously it was
362 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
363 are unsuable for non-message-specific reasons - eg. connection timeout,
364 banner-time rejection.
366 JH/16 Fix logging of delivery remote port, when specified by router, under
369 PP/06 Repair manualroute's ability to take options in any order, even if one
370 is the name of a transport.
373 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
375 JH/17 Change the list-building routines interface to use the expanding-string
376 triplet model, for better allocation and copying behaviour.
378 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
379 Previously it was constructed the first time a possibly-matching string
380 was met in the configuration file input during startup; now it is done
383 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
384 compatible one, to avoid the (poorly documented) possibility of a config
385 file in the working directory redirecting the DB files, possibly correpting
386 some existing file. CVE-2017-10140 assigned for BDB.
388 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
389 cache-hot. Previously, although the result was properly cached, the
390 initial verify call returned a defer.
392 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
393 the main verify for receipient in uncached-mode.
395 JH/22 Retire historical build files to an "unsupported" subdir. These are
396 defined as "ones for which we have no current evidence of testing".
398 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
399 if present. Previously it was ignored.
401 JH/24 Start using specified-initialisers in C structure init coding. This is
402 a C99 feature (it's 2017, so now considered safe).
404 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
405 if was a fixed-sized field and bitmask ops via macros; it is now more
408 PP/07 GitHub PR 56: Apply MariaDB build fix.
409 Patch provided by Jaroslav Škarvada.
411 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
412 during Coverity cleanups [4.87 JH/47]
413 Diagnosis and fix provided by Michael Fischer v. Mollard.
415 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
416 the right size to place the terminating semicolon on its own folded
417 line, the header hash was calculated to an incorrect value thanks to
418 the (relaxed) space the fold became.
420 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
421 and confused the parent.
423 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
424 which could crash as a result. This could lead to undeliverable messages.
426 JH/28 Logging: "next input sent too soon" now shows where input was truncated
429 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
430 matters on fast-turnover and PID-randomising systems, which were getting
431 out-of-order delivery.
433 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
434 a possibly-overlapping copy. The symptom was that "Remote host closed
435 connection in response to HELO" was logged instead of the actual 4xx
438 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
439 Previously only that bufferd was discarded, resulting in SYMTP command
442 JH/32 DKIM: when a message has multiple signatures matching an identity given
443 in dkim_verify_signers, run the dkim acl once for each. Previously only
444 one run was done. Bug 2189.
446 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
447 "panic the current process" to "deliberately defer". The panic log is
448 still written with the problem list name; the mail and reject logs now
449 get a temp-reject line for the message that was being handled, saying
450 something like "domains check lookup or other defer". The SMTP 451
451 message is still "Temporary local problem".
453 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
454 A crafted sequence of BDAT commands could result in in-use memory beeing
455 freed. CVE-2017-16943.
457 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
458 from SMTP input. Previously it was always done; now only done for DATA
459 and not BDAT commands. CVE-2017-16944.
461 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
462 to the message (such as an overlong header line). Previously this was
463 not done and we did not exit BDAT mode. Followon from the previous item
464 though a different problem.
470 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
471 than -2003 did; needs libidn2 in addition to libidn.
473 JH/02 The path option on a pipe transport is now expanded before use.
475 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
476 Patch provided by "Björn", documentation fix added too.
478 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
479 missing a wire-to-host endian conversion.
481 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
482 close after a BDAT command line could be taken as a following command,
483 giving a synch failure. Fix by only checking for synch immediately
484 before acknowledging the chunk.
486 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
487 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
488 macro. Patches provided by Josh Soref.
490 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
491 Previously we did not; the RFC seems ambiguous and VRFY is not listed
492 by IANA as a service extension. However, John Klensin suggests that we
495 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
496 the dkim code may be unix-mode line endings rather than smtp wire-format
497 CRLF, so prepend a CR to any bare LF.
499 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
500 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
502 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
503 the first were themselves being wrongly included in the feed into dkim
504 processing; with most chunk sizes in use this resulted in an incorrect
505 body hash calculated value.
507 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
508 DKIM signature block, for verification. Although advised against by
509 standards it is specifically not ruled illegal.
511 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
513 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
514 missing a body hash (the bh= tag).
516 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
517 It seems that HAProxy sends the Proxy Protocol information in clear and
518 only then does a TLS startup, so do the same.
520 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
521 TCP connections (such as for Spamd) unless the daemon successfully set
522 Fast Open mode on its listening sockets. This fixes breakage seen on
523 too-old kernels or those not configured for Fast Open, at the cost of
524 requiring both directions being enabled for TFO, and TFO never being used
525 by non-daemon-related Exim processes.
527 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
528 endings, at least on the first header line. Try to canonify any that get
529 past that check, despite the cost.
531 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
532 now limited to an arbitrary five deep, while parsing addresses with the
533 strip_excess_angle_brackets option enabled.
535 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
536 instead leave the unprompted TLS handshake in socket buffer for the
537 TLS library to consume.
539 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
541 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
543 JH/16 Drop variables when they go out of scope. Memory management drops a whole
544 region in one operation, for speed, and this leaves assigned pointers
545 dangling. Add checks run only under the testsuite which checks all
546 variables at a store-reset and panics on a dangling pointer; add code
547 explicitly nulling out all the variables discovered. Fixes one known
548 bug: a transport crash, where a dangling pointer for $sending_ip_address
549 originally assigned in a verify callout, is re-used.
551 PP/06 Drop '.' from @INC in various Perl scripts.
553 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
555 PP/08 Reduce a number of compilation warnings under clang; building with
556 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
557 should be warning-free.
559 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
561 HS/01 Fix portability problems introduced by PP/08 for platforms where
562 realloc(NULL) is not equivalent to malloc() [SunOS et al].
564 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
565 chunk. This allows us to accept broken chunked messages. We need a more
566 general solution here.
568 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
569 already-broken messages in the queue.
571 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
573 JH/19 Fix reference counting bug in routing-generated-address tracking.
579 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
580 supports it and a size is available (ie. the sending peer gave us one).
582 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
583 years of being deprecated). The replacements are the ACLs
584 acl_smtp_mime and acl_not_smtp_mime.
586 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
587 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
588 or even an in-clear connection were permitted. Now, if the host lookup
589 was dnssec and dane was requested then the host is only used if the
590 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
591 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
592 if one fails this test.
593 This means that a poorly-configured remote DNS will make it incommunicado;
594 but it protects against a DNS-interception attack on it.
596 JH/04 Bug 1810: make continued-use of an open smtp transport connection
597 non-noisy when a race steals the message being considered.
599 JH/05 If main configuration option tls_certificate is unset, generate a
600 self-signed certificate for inbound TLS connections.
602 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
603 in rewrites and routers.
605 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
606 and logged a warning sing 4.83; now they are a configuration file error.
608 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
609 (lacking @domain). Apply the same qualification processing as RCPT.
611 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
613 JH/10 Support ${sha256:} applied to a string (as well as the previous
616 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
617 a cutthrough deliver is pending, as we always want to make a connection.
618 This also avoids re-routing the message when later placing the cutthrough
619 connection after a verify cache hit.
620 Do not update it with the verify result either.
622 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
623 when routing results in more than one destination address.
625 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
626 signing (which inhibits the cutthrough capability). Previously only
627 the presence of an option was tested; now an expansion evaluating as
628 empty is permissible (obviously it should depend only on data available
629 when the cutthrough connection is made).
631 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
632 the relevant preceding SMTP command did not note the pipelining mode.
634 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
635 Previously they were not counted.
637 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
638 as one having no matching records. Previously we deferred the message
639 that needed the lookup.
641 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
642 distinguished as "(=".
644 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
645 for missing MX records. Previously it only worked for missing A records.
647 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
649 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
650 after the data-go-ahead and data-ack. Patch from Jason Betts.
652 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
653 even for a "none" policy. Patch from Tony Meyer.
655 JH/22 Fix continued use of a connection for further deliveries. If a port was
656 specified by a router, it must also match for the delivery to be
659 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
660 When one of the recipients of a message was unsuitable for the connection
661 (has no matching addresses), we lost track of needing to mark it
662 deferred. As a result mail would be lost.
664 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
666 JH/25 Decoding ACL controls is now done using a binary search; the source code
667 takes up less space and should be simpler to maintain. Merge the ACL
668 condition decode tables also, with similar effect.
670 JH/26 Fix problem with one_time used on a redirect router which returned the
671 parent address unchanged. A retry would see the parent address marked as
672 delivered, so not attempt the (identical) child. As a result mail would
675 JH/27 Fix a possible security hole, wherein a process operating with the Exim
676 UID can gain a root shell. Credit to http://www.halfdog.net/ for
677 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
680 JH/28 Enable {spool,log} filesystem space and inode checks as default.
681 Main config options check_{log,spool}_{inodes,space} are now
682 100 inodes, 10MB unless set otherwise in the configuration.
684 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
685 Previously it only applied to the main-section connection policy
688 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
690 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
691 by me. Added RFC7919 DH primes as an alternative.
693 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
694 are not in the system include path.
696 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
697 GnuTLS, when a session startup failed (eg because the client disconnected)
698 Exim did stdio operations after fclose. This was exposed by a recent
699 change which nulled out the file handle after the fclose.
701 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
702 signed directly by the cert-signing cert, rather than an intermediate
703 OCSP-signing cert. This is the model used by LetsEncrypt.
705 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
707 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
708 an incoming connection.
710 HS/02 Bug 1802: Do not half-close the connection after sending a request
713 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
714 fallback to "prime256v1".
716 JH/34 SECURITY: Use proper copy of DATA command in error message.
717 Could leak key material. Remotely exploitable. CVE-2016-9963.
723 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
724 and 3.4.4 - once the server is enabled to respond to an OCSP request
725 it does even when not requested, resulting in a stapling non-aware
726 client dropping the TLS connection.
728 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
729 support variable-length bit vectors. No functional change.
731 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
732 The I= interface field on outgoing lines is now after the H= remote
733 host field, same as incoming lines. There is a separate
734 outgoing_interface log selector which allows you to disable the
737 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
738 If not running log_selector +smtp_connection the mainlog would be held
739 open indefinitely after a "too many connections" event, including to a
740 deleted file after a log rotate. Leave the per net connection logging
741 leaving it open for efficiency as that will be quickly detected by the
742 check on the next write.
744 HS/01 Bug 1671: Fix post transport crash.
745 Processing the wait-<transport> messages could crash the delivery
746 process if the message IDs didn't exist for some reason. When
747 using 'split_spool_directory=yes' the construction of the spool
748 file name failed already, exposing the same netto behaviour.
750 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
751 mime_regex ACL conditions.
753 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
754 to DSN fail messages (bounces): remote IP, remote greeting, remote response
755 to HELO, local diagnostic string.
757 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
758 log line to debug. Even when configured with a tls authenticator many
759 client connections are expected to not authenticate in this way, so
760 an authenticate fail is not an error.
762 HS/02 Add the Exim version string to the process info. This way exiwhat
763 gives some more detail about the running daemon.
765 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
766 matter for fast-change records such as DNSBLs.
768 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
769 retry record, even if constant. There may be multiple transports with
770 different interface settings and the retry behaviour needs to be kept
773 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
775 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
777 JH/10 Bug 840: fix log_defer_output option of pipe transport
779 JH/11 Bug 830: use same host for all RCPTS of a message, even under
780 hosts_randomize. This matters a lot when combined with mua_wrapper.
782 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
783 ${quote_pgsql:<string>} operator.
785 JH/13 Bug 1708: avoid misaligned access in cached lookup.
787 JH/14 Change header file name for freeradius-client. Relevant if compiling
788 with Radius support; from the Gentoo tree and checked under Fedora.
790 JH/15 Bug 1712: Introduce $prdr_requested flag variable
792 JH/16 Bug 1714: Permit an empty string as expansion result for transport
793 option transport_filter, meaning no filtering.
795 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
797 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
798 defaults to "*" (all hosts). The variable is now available when not built
799 with TLS, default unset, mainly to enable keeping the testsuite sane.
800 If a server certificate is not supplied (via tls_certificate) an error is
801 logged, and clients will find TLS connections fail on startup. Presumably
802 they will retry in-clear.
803 Packagers of Exim are strongly encouraged to create a server certificate
804 at installation time.
806 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
807 with the $config_file variable.
809 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
810 in transport context, after the attempt, and per-recipient. The latter type
811 is per host attempted. The event data is the error message, and the errno
812 information encodes the lookup type (A vs. MX) used for the (first) host,
813 and the trailing two digits of the smtp 4xx response.
815 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
816 to write to mainlog (or rejectlog, paniclog) in the window between file
817 creation and permissions/ownership being changed. Particularly affects
818 installations where exicyclog is run as root, rather than exim user;
819 result is that the running daemon panics and dies.
821 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
823 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
824 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
825 "pri" and "weight". Note that the previous implicit priority given by the
826 list order is no longer honoured.
828 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
831 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
832 by defining SUPPORT_SOCKS.
834 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
835 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
836 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
837 variables are renamed to proxy_{local,external}_{address,port}.
839 JH/27 Move Internationalisation support from Experimental to mainline, enabled
840 for a build by defining SUPPORT_I18N
842 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
843 of the query string, and make ${quote_redis:} do that quoting.
845 JH/29 Move Events support from Experimental to mainline, enabled by default
846 and removable for a build by defining DISABLE_EVENT.
848 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
850 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
851 cached by the daemon.
853 JH/32 Move Redis support from Experimental to mainline, enabled for a build
854 by defining LOOKUP_REDIS. The libhiredis library is required.
856 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
857 keys are given for lookup.
859 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
860 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
861 only supported when built with TLS support. The PolarSSL SHA routines
862 are still used when the TLS library is too old for convenient support.
864 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
865 openssl_options), for security. OpenSSL forces this from version 1.1.0
866 server-side so match that on older versions.
868 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
869 allocation for $value could be released as the expansion processing
870 concluded, but leaving the global pointer active for it.
872 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
873 and to use the domains and local_parts ACL conditions.
875 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
876 incorrectly not doubled on cutthrough transmission, hence seen as a
877 body-termination at the receiving system - resulting in truncated mails.
878 Commonly the sender saw a TCP-level error, and retransmitted the message
879 via the normal store-and-forward channel. This could result in duplicates
880 received - but deduplicating mailstores were liable to retain only the
881 initial truncated version.
883 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
885 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
887 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
888 we're in there, support oversigning also; bug 1309.
890 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
892 HS/04 Add support for keep_environment and add_environment options.
894 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
895 either intentional arithmetic overflow during PRNG, or testing config-
898 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
899 delivery resulted in actual delivery. Cancel cutthrough before DATA
902 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
903 rejects a recipient: pass the reject to the originator.
905 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
906 Many were false-positives and ignorable, but it's worth fixing the
909 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
910 for the new environment-manipulation done at startup. Move the routines
911 from being local to tls.c to being global via the os.c file.
913 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
914 an extract embedded as result-arg for a map, the first arg for extract
915 is unavailable so we cannot tell if this is a numbered or keyed
916 extraction. Accept either.
922 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
925 JH/02 The smtp transport option "multi_domain" is now expanded.
927 JH/03 The smtp transport now requests PRDR by default, if the server offers
930 JH/04 Certificate name checking on server certificates, when exim is a client,
931 is now done by default. The transport option tls_verify_cert_hostnames
932 can be used to disable this per-host. The build option
933 EXPERIMENTAL_CERTNAMES is withdrawn.
935 JH/05 The value of the tls_verify_certificates smtp transport and main options
936 default to the word "system" to access the system default CA bundle.
937 For GnuTLS, only version 3.0.20 or later.
939 JH/06 Verification of the server certificate for a TLS connection is now tried
940 (but not required) by default. The verification status is now logged by
941 default, for both outbound TLS and client-certificate supplying inbound
944 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
947 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
948 Status Notification (bounce) messages are now MIME format per RFC 3464.
949 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
950 under the control of the dsn_advertise_hosts option, and routers may
951 have a dsn_lasthop option.
953 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
954 default, modifiable by a malware= option. The list separator for
955 the options can now be changed in the usual way. Bug 68.
957 JH/10 The smtp_receive_timeout main option is now expanded before use.
959 JH/11 The incoming_interface log option now also enables logging of the
960 local interface on delivery outgoing connections.
962 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
963 if the interface and destination host and port all match.
965 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
968 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
969 Patch from Andrew Lewis.
971 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
972 now supports optional time-restrictions, weighting, and priority
973 modifiers per server. Patch originally by <rommer@active.by>.
975 JH/16 The spamd_address main option now supports a mixed list of local
976 and remote servers. Remote servers can be IPv6 addresses, and
977 specify a port-range.
979 JH/17 Bug 68: The spamd_address main option now supports an optional
980 timeout value per server.
982 JH/18 Bug 1581: Router and transport options headers_add/remove can
983 now have the list separator specified.
985 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
988 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
991 JH/21 Support for the A6 type of dns record is withdrawn.
993 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
994 rather than the verbs used.
996 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
997 from 255 to 1024 chars.
999 JH/24 Verification callouts now attempt to use TLS by default.
1001 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1002 are generic router options now. The defaults didn't change.
1004 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1005 Original patch from Alexander Shikoff, worked over by JH.
1007 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1008 files if ZCAT_COMMAND is not executable.
1010 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1012 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1014 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1015 Normally benign, it bites when the pair was led to by a CNAME;
1016 modern usage is to not canonicalize the domain to a CNAME target
1017 (and we were inconsistent anyway for A-only vs AAAA+A).
1019 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1021 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1022 when evaluating $sender_host_dnssec.
1024 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1025 $sender_helo_dnssec variable.
1027 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1029 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1031 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1033 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1034 documented as working, but never had. Support all but $spam_report.
1036 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1037 added for tls authenticator.
1039 HS/03 Add perl_taintmode main config option
1045 TL/01 When running the test suite, the README says that variables such as
1046 no_msglog_check are global and can be placed anywhere in a specific
1047 test's script, however it was observed that placement needed to be near
1048 the beginning for it to behave that way. Changed the runtest perl
1049 script to read through the entire script once to detect and set these
1050 variables, reset to the beginning of the script, and then run through
1051 the script parsing/test process like normal.
1053 TL/02 The BSD's have an arc4random API. One of the functions to induce
1054 adding randomness was arc4random_stir(), but it has been removed in
1055 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1056 function when detected.
1058 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1059 cause callback expansion.
1061 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1062 syntax errors in an expansion can be treated as a string instead of
1063 logging or causing an error, due to the internal use of bool_lax
1064 instead of bool when processing it.
1066 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1067 server certificates when making smtp deliveries.
1069 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1071 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1073 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1075 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1076 Merged patch from Sebastian Wiedenroth.
1078 JH/05 Fix results-pipe from transport process. Several recipients, combined
1079 with certificate use, exposed issues where response data items split
1080 over buffer boundaries were not parsed properly. This eventually
1081 resulted in duplicates being sent. This issue only became common enough
1082 to notice due to the introduction of connection certificate information,
1083 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1085 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1086 size buffer was used, resulting in syntax errors when an expansion
1089 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1090 version 3.3.6 or later.
1092 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1093 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1094 both become "event_action", the variables become $event_name, $event_data
1095 and $event_defer_errno. There is a new variable $verify_mode, usable in
1096 routers, transports and related events. The tls:cert event is now also
1097 raised for inbound connections, if the main configuration event_action
1100 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1101 early OCSP support, but no stapling (appears to be less than 1.0.0).
1103 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1104 server certificate names available under the smtp transport option
1105 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1108 JH/10 Time-related extraction expansions from certificates now use the main
1109 option "timezone" setting for output formatting, and are consistent
1110 between OpenSSL and GnuTLS compilations. Bug 1541.
1112 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1113 encoded parameter in the incoming message. Bug 1558.
1115 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1116 include certificate info, eximon was claiming there were spoolfile
1119 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1121 JH/14 Log delivery-related information more consistently, using the sequence
1122 "H=<name> [<ip>]" wherever possible.
1124 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1125 are problematic for Debian distribution, omit them from the release
1128 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1130 JH/16 Fix string representation of time values on 64bit time_t architectures.
1133 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1134 output list separator was used.
1139 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1140 checkers that were complaining about end of non-void function with no
1143 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1144 This was a regression introduced in 4.83 by another bugfix.
1146 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1148 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1149 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1155 TF/01 Correctly close the server side of TLS when forking for delivery.
1157 When a message was received over SMTP with TLS, Exim failed to clear up
1158 the incoming connection properly after forking off the child process to
1159 deliver the message. In some situations the subsequent outgoing
1160 delivery connection happened to have the same fd number as the incoming
1161 connection previously had. Exim would try to use TLS and fail, logging
1162 a "Bad file descriptor" error.
1164 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1165 utilities have not been installed.
1167 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1168 temporary space as the ACL may create new global variables.
1170 TL/01 LDAP support uses per connection or global context settings, depending
1171 upon the detected version of the libraries at build time.
1173 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1174 to extract and use the src ip:port in logging and expansions as if it
1175 were a direct connection from the outside internet. PPv2 support was
1176 updated based on HAProxy spec change in May 2014.
1178 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1180 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1181 Properly escape header and check for NULL return.
1183 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1186 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1188 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1189 characters in header names, implemented as a verify condition.
1190 Contributed by Michael Fischer v. Mollard.
1192 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1193 results permerror and temperror. Previous values are deprecated but
1194 still accepted. In a future release, err_perm and err_temp will be
1195 completely removed, which will be a backward incompatibility if the
1196 ACL tests for either of these two old results. Patch contributed by
1197 user bes-internal on the mailing list.
1199 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1201 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1202 selectors, in both main and reject logs.
1204 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1207 JH/07 Add malware type "sock" for talking to simple daemon.
1209 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1211 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1212 routers/transports under cutthrough routing.
1214 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1215 numbers. Touch up "bool" conditional to keep the same definition.
1217 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1219 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1221 JH/12 Expand items in router/transport headers_add or headers_remove lists
1222 individually rather than the list as a whole. Bug 1452.
1224 Required for reasonable handling of multiple headers_ options when
1225 they may be empty; requires that headers_remove items with embedded
1226 colons must have them doubled (or the list-separator changed).
1228 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1229 view the policy declared in the DMARC record. Currently, $dmarc_status
1230 is a combined value of both the record presence and the result of the
1233 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1235 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1236 dnslookup router and the smtp transport (applying to the forward
1239 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1240 of ldap servers used for a specific lookup. Patch provided by Heiko
1243 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1244 New variable $lookup_dnssec_authenticated for observability.
1246 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1247 Patch submitted by Lars Timman.
1249 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1251 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1252 Requires trusted mode and valid format message id, aborts otherwise.
1253 Patch contributed by Heiko Schlichting.
1255 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1256 certextract with support for various fields. Bug 1358.
1258 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1259 is requested by default, modifiable by smtp transport option
1262 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1263 operate on certificate variables to give certificate fingerprints
1264 Also new ${sha256:cert_variable}.
1266 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1268 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1269 Christian Aistleitner.
1271 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1273 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1274 file. Patch from Wolfgang Breyha.
1276 JH/25 Expand the coverage of the delivery $host and $host_address to
1277 client authenticators run in verify callout. Bug 1476.
1279 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1280 align with daemon_smtp_ports. Bug 72.
1282 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1283 support and error reporting did not work properly.
1285 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1286 and is readable. Patch from Andrew Colin Kissa.
1288 TL/14 Enhance documentation of ${run expansion and how it parses the
1289 commandline after expansion, particularly in the case when an
1290 unquoted variable expansion results in an empty value.
1292 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1294 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1295 by renaming away from it. Addresses GH issue 15, reported by
1298 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1300 TL/15 SECURITY: prevent double expansion in math comparison functions
1301 (can expand unsanitized data). Not remotely exploitable.
1308 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1310 PP/02 Make -n do something, by making it not do something.
1311 When combined with -bP, the name of an option is not output.
1313 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1316 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1317 $sender_host_name and config options to manage this, and basic check
1320 PP/05 DSCP support for outbound connections and control modifier for inbound.
1322 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1323 (Only plugin which currently uses this is kerberos4, which nobody should
1324 be using, but we should make it available and other future plugins might
1325 conceivably use it, even though it would break NAT; stuff *should* be
1326 using channel bindings instead).
1328 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1329 name; added for Sendmail compatibility; requires admin caller.
1330 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1331 just ignore it); requires trusted caller.
1332 Also parse but ignore: -Ac -Am -X<logfile>
1335 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1337 TL/02 Add +smtp_confirmation as a default logging option.
1339 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1340 Patch by Magnus Holmgren from 2007-02-20.
1342 TL/04 Bugzilla 1281 - Spec typo.
1343 Bugzilla 1283 - Spec typo.
1344 Bugzilla 1290 - Spec grammar fixes.
1346 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1348 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1350 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1351 mailing list by by Dmitry Isaikin.
1353 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1355 JH/02 Support "G" suffix to numbers in ${if comparisons.
1357 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1359 NM/01 Bugzilla 1197 - Spec typo
1360 Bugzilla 1196 - Spec examples corrections
1362 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1364 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1365 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1368 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1369 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1371 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1372 "acl {{name}{arg}...}", and optional args on acl condition
1375 JH/05 Permit multiple router/transport headers_add/remove lines.
1377 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1379 JH/07 Avoid using a waiting database for a single-message-only transport.
1380 Performance patch from Paul Fisher. Bugzilla 1262.
1382 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1385 JH/09 Add $headers_added variable, with content from use of ACL modifier
1386 add_header (but not yet added to the message). Bugzilla 199.
1388 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1389 Pulled from Bugzilla 817 by Wolfgang Breyha.
1391 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1393 (nb: this is the same fix as in Exim 4.80.1)
1395 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1398 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1401 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1403 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1404 advertise SMTP AUTH mechanism to us, instead of a generic
1405 protocol violation error. Also, make Exim more robust to bad
1406 data from the Dovecot auth socket.
1408 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1410 When a queue runner is handling a message, Exim first routes the
1411 recipient addresses, during which it prunes them based on the retry
1412 hints database. After that it attempts to deliver the message to
1413 any remaining recipients. It then updates the hints database using
1416 So if a recipient address works intermittently, it can get repeatedly
1417 deferred at routing time. The retry hints record remains fresh so the
1418 address never reaches the final cutoff time.
1420 This is a fairly common occurrence when a user is bumping up against
1421 their storage quota. Exim had some logic in its local delivery code
1422 to deal with this. However it did not apply to per-recipient defers
1423 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1425 This change adds a proper retry rule check during routing so that the
1426 final cutoff time is checked against the message's age. We only do
1427 this check if there is an address retry record and there is not a
1428 domain retry record; this implies that previous attempts to handle
1429 the address had the retry_use_local_parts option turned on. We use
1430 this as an approximation for the destination being like a local
1431 delivery, as in LMTP.
1433 I suspect this new check makes the old local delivery cutoff check
1434 redundant, but I have not verified this so I left the code in place.
1436 TF/02 Correct gecos expansion when From: is a prefix of the username.
1438 Test 0254 submits a message to Exim with the header
1442 When I ran the test suite under the user fanf2, Exim expanded
1443 the header to contain my full name, whereas it should have added
1444 a Resent-Sender: header. It erroneously treats any prefix of the
1445 username as equal to the username.
1447 This change corrects that bug.
1449 GF/01 DCC debug and logging tidyup
1450 Error conditions log to paniclog rather than rejectlog.
1451 Debug lines prefixed by "DCC: " to remove any ambiguity.
1453 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1455 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1456 Bug spotted by Jeremy Harris; was flawed since initial commit.
1457 Would have resulted in OCSP responses post-SNI triggering an Exim
1458 NULL dereference and crash.
1460 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1462 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1463 Bug detection, analysis and fix by Samuel Thibault.
1464 Bugzilla 1331, Debian bug #698092.
1466 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1468 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1469 Server implementation by Todd Lyons, client by JH.
1470 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1471 config variable "prdr_enable" controls whether the server
1472 advertises the facility. If the client requests PRDR a new
1473 acl_data_smtp_prdr ACL is called once for each recipient, after
1474 the body content is received and before the acl_smtp_data ACL.
1475 The client is controlled by both of: a hosts_try_prdr option
1476 on the smtp transport, and the server advertisement.
1477 Default client logging of deliveries and rejections involving
1478 PRDR are flagged with the string "PRDR".
1480 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1481 fclose(). Diagnosis by Todd Lyons.
1483 PP/17 Update configure.default to handle IPv6 localhost better.
1484 Patch by Alain Williams (plus minor tweaks).
1487 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1488 This is now consistent with GnuTLS, and is now documented: the
1489 previous undocumented portable approach to treating the option as
1490 unset was to force an expansion failure. That still works, and
1491 an empty string is now equivalent.
1493 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1494 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1495 not performing validation itself.
1497 PP/20 Added force_command boolean option to pipe transport.
1498 Patch from Nick Koston, of cPanel Inc.
1500 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1503 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1505 PP/21 Fix eximon continuous updating with timestamped log-files.
1506 Broken in a format-string cleanup in 4.80, missed when I repaired the
1507 other false fix of the same issue.
1508 Report and fix from Heiko Schlichting.
1511 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1512 Report from Prashanth Katuri.
1514 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1515 It's SecureTransport, so affects any MacOS clients which use the
1516 system-integrated TLS libraries, including email clients.
1518 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1519 using a MIME ACL for non-SMTP local injection.
1520 Report and assistance in diagnosis by Warren Baker.
1522 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1524 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1526 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1527 last id that failed so it may be referenced in subsequent ACL's.
1529 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1532 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1533 ldap library initialization, allowing self-signed CA's to be
1534 used. Also properly sets require_cert option later in code by
1535 using NULL (global ldap config) instead of ldap handle (per
1536 session). Bug diagnosis and testing by alxgomz.
1538 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1539 the src/util/ subdirectory.
1541 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1542 renamed to Transport Post Delivery Action by Jeremy Harris, as
1545 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1546 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1547 redis_servers = needs to be configured which will be used by the redis
1548 lookup. Patch from Warren Baker, of The Packet Hub.
1550 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1552 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1553 hostname or reverse DNS when processing a host list. Used suggestions
1554 from multiple comments on this bug.
1556 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1558 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1559 Missed a few lines, added it to make the runtest require no keyboard
1562 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1563 contains upper case chars. Make router use caseful_local_part.
1565 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1566 support when GnuTLS has been built with p11-kit.
1572 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1574 This, or similar/improved, will also be change PP/11 of 4.82.
1580 PP/01 Handle short writes when writing local log-files.
1581 In practice, only affects FreeBSD (8 onwards).
1582 Bugzilla 1053, with thanks to Dmitry Isaikin.
1584 NM/01 Bugzilla 949 - Documentation tweak
1586 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1589 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1591 PP/02 Implemented gsasl authenticator.
1593 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1595 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1596 `pkg-config foo` for cflags/libs.
1598 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1599 with rest of GSASL and with heimdal_gssapi.
1601 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1602 `pkg-config foo` for cflags/libs for the TLS implementation.
1604 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1605 properties get this fed in as external SSF. A number of robustness
1606 and debugging improvements to the cyrus_sasl authenticator.
1608 PP/08 cyrus_sasl server now expands the server_realm option.
1610 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1611 Patch by Jeremy Harris.
1613 PP/10 Added dbmjz lookup type.
1615 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1617 PP/12 MAIL args handles TAB as well as SP, for better interop with
1618 non-compliant senders.
1619 Analysis and variant patch by Todd Lyons.
1621 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1622 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1623 Patch from Dirk Mueller <dmueller@suse.com>
1625 PP/13 tls_peerdn now print-escaped for spool files.
1626 Observed some $tls_peerdn in wild which contained \n, which resulted
1627 in spool file corruption.
1629 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1630 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1631 or write after TLS renegotiation, which otherwise led to messages
1634 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1635 as a tracking header (ie: a signed header comes before the signature).
1636 Patch from Wolfgang Breyha.
1638 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1639 comma-sep list; embedded commas doubled.
1641 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1643 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1645 Report and patch from Dmitry Banschikov.
1647 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1648 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1649 failures. We appear to now support TLS1.1+ with Exim.
1651 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1652 lets Exim select keys and certificates based upon TLS SNI from client.
1653 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1654 before an outbound SMTP session. New log_selector, +tls_sni.
1656 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1657 NULL dereference. Report and patch from Alun Jones.
1659 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1660 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1661 Not seeing resolver debug output on NetBSD, but suspect this is a
1662 resolver implementation change.
1664 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1665 Left warnings. Added "eximon gdb" invocation mode.
1667 PP/21 Defaulting "accept_8bitmime" to true, not false.
1669 PP/22 Added -bw for inetd wait mode support.
1671 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1672 locate the relevant includes and libraries. Made this the default.
1674 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1675 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1677 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1678 This may cause build issues on older platforms.
1680 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1681 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1682 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1683 Added SNI support via GnuTLS too.
1684 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1686 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1688 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1689 Applied second patch from Janne, implementing suggestion to default
1690 multiple-strings-in-record handling to match SPF spec.
1692 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1694 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1695 read-only, out of scope).
1696 Patch from Wolfgang Breyha, report from Stuart Northfield.
1698 PP/29 Fix three issues highlighted by clang analyser static analysis.
1699 Only crash-plausible issue would require the Cambridge-specific
1700 iplookup router and a misconfiguration.
1701 Report from Marcin Mirosław.
1703 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1705 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1706 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1707 As part of this, removing so much warning spew let me fix some minor
1708 real issues in debug logging.
1710 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1711 assignment on my part. Fixed.
1713 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1714 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1715 Janne Snabb (who went above and beyond: thank you).
1717 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1718 string otherwise requires a connection and a bunch more work and it's
1719 relatively easy to get wrong. Should also expose TLS library linkage
1722 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1723 64-bit ${eval} (JH/03).
1725 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1726 GNU libc to support some of the 64-bit stuff, should not lead to
1727 conflicts. Defined before os.h is pulled in, so if a given platform
1728 needs to override this, it can.
1730 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1731 protection layer was required, which is not implemented.
1732 Bugzilla 1254, patch from Wolfgang Breyha.
1734 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1735 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1736 tls_dhparam take prime identifiers. Also unbreak combination of
1737 OpenSSL+DH_params+TLSSNI.
1739 PP/39 Disable SSLv2 by default in OpenSSL support.
1745 PP/01 Solaris build fix for Oracle's LDAP libraries.
1746 Bugzilla 1109, patch from Stephen Usher.
1748 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1750 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1753 TF/02 Fix a couple more cases where we did not log the error message
1754 when unlink() failed. See also change 4.74-TF/03.
1756 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1757 lock up or crash if it happened to be inside a call to libc when it
1758 got a SIGUSR1 from exiwhat.
1760 The SIGUSR1 handler appends the current process status to the process
1761 log which is later printed by exiwhat. It used to use the general
1762 purpose logging code to do this, but several functions it calls are
1763 not safe for signals.
1765 The new output code in the SIGUSR1 handler is specific to the process
1766 log, and simple enough that it's easy to inspect for signal safety.
1767 Removing some special cases also simplifies the general logging code.
1768 Removing the spurious timestamps from the process log simplifies
1771 TF/04 Improved ratelimit ACL condition.
1773 The /noupdate option has been deprecated in favour of /readonly which
1774 has clearer semantics. The /leaky, /strict, and /readonly update modes
1775 are mutually exclusive. The update mode is no longer included in the
1776 database key; it just determines when the database is updated. (This
1777 means that when you upgrade Exim will forget old rate measurements.)
1779 Exim now checks that the per_* options are used with an update mode that
1780 makes sense for the current ACL. For example, when Exim is processing a
1781 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1782 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1783 must specify per_mail/readonly. If you omit the update mode it defaults to
1784 /leaky where that makes sense (as before) or /readonly where required.
1786 The /noupdate option is now undocumented but still supported for
1787 backwards compatibility. It is equivalent to /readonly except that in
1788 ACLs where /readonly is required you may specify /leaky/noupdate or
1789 /strict/noupdate which are treated the same as /readonly.
1791 A useful new feature is the /count= option. This is a generalization
1792 of the per_byte option, so that you can measure the throughput of other
1793 aggregate values. For example, the per_byte option is now equivalent
1794 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1796 The per_rcpt option has been generalized using the /count= mechanism
1797 (though it's more complicated than the per_byte equivalence). When it is
1798 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1799 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1800 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1801 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1802 Note that using per_rcpt with a non-readonly update mode in more than
1803 one ACL will cause the recipients to be double-counted. (The per_mail
1804 and per_byte options don't have this problem.)
1806 The handling of very low rates has changed slightly. If the computed rate
1807 is less than the event's count (usually one) then this event is the first
1808 after a long gap. In this case the rate is set to the same as this event's
1809 count, so that the first message of a spam run is counted properly.
1811 The major new feature is a mechanism for counting the rate of unique
1812 events. The new per_addr option counts the number of different
1813 recipients that someone has sent messages to in the last time period. It
1814 behaves like per_rcpt if all the recipient addresses are different, but
1815 duplicate recipient addresses do not increase the measured rate. Like
1816 the /count= option this is a general mechanism, so the per_addr option
1817 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1818 example, measure the rate that a client uses different sender addresses
1819 with the options per_mail/unique=$sender_address. There are further
1820 details in the main documentation.
1822 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1824 TF/06 Removed a few PCRE remnants.
1826 TF/07 Automatically extract Exim's version number from tags in the git
1827 repository when doing development or release builds.
1829 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1830 Bugzilla 879. Patch from Paul Fisher.
1832 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1833 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1836 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1838 PP/05 Variable $av_failed, true if the AV scanner deferred.
1839 Bugzilla 1078. Patch from John Horne.
1841 PP/06 Stop make process more reliably on build failure.
1842 Bugzilla 1087. Patch from Heiko Schlittermann.
1844 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1845 Bugzilla 1089. Patch from Heiko Schlittermann.
1847 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1848 Bugzilla 1131. Patch from Holger Weiß.
1850 PP/09 Handle IPv6 addresses with SPF.
1851 Bugzilla 860. Patch from Wolfgang Breyha.
1853 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1855 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1858 PP/11 match_* no longer expand right-hand-side by default.
1859 New compile-time build option, EXPAND_LISTMATCH_RHS.
1860 New expansion conditions, "inlist", "inlisti".
1862 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1864 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1866 PP/14 fix log_write() format string regression from TF/03.
1867 Bugzilla 1152. Patch from Dmitry Isaikin.
1873 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1875 PP/02 Harmonised TLS library version reporting; only show if debugging.
1876 Layout now matches that introduced for other libraries in 4.74 PP/03.
1878 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1880 PP/04 New "dns_use_edns0" global option.
1882 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1885 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1886 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1888 TK/01 Updated PolarSSL code to 0.14.2.
1889 Bugzilla 1097. Patch from Andreas Metzler.
1891 PP/07 Catch divide-by-zero in ${eval:...}.
1892 Fixes bugzilla 1102.
1894 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1897 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1898 format-string attack -- SECURITY: remote arbitrary code execution.
1900 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1901 time unintentionally subject to list matching rules, letting the header
1902 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1903 arbitrary string expansion). This allowed for information disclosure.
1905 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1906 INT_MIN/-1 -- value coerced to INT_MAX.
1912 NM/01 Workaround for PCRE version dependency in version reporting
1915 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1916 This fixes portability to compilers other than gcc, notably
1917 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1919 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1920 makefiles for portability to HP-UX and POSIX correctness.
1922 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1923 Also via indented variable definition in the Makefile.
1924 (Debugging by Oliver Heesakkers).
1926 PP/02 Restore caching of spamd results with expanded spamd_address.
1927 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1929 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1930 Improves build reliability. Fix from: Frank Elsner
1932 NM/02 Fix wide character breakage in the rfc2047 coding
1933 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1935 NM/03 Allow underscore in dnslist lookups
1936 Fixes bug 1026. Patch from Graeme Fowler
1938 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1939 Code patches from Adam Ciarcinski of NetBSD.
1941 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1944 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1945 is logged, to avoid truncation. Patch from John Horne.
1947 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1948 Patch from Jakob Hirsch.
1950 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1951 SQL string expansion failure details.
1952 Patch from Andrey Oktyabrski.
1954 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1955 Patch from Simon Arlott.
1957 PP/09 New lookups functionality failed to compile on old gcc which rejects
1958 extern declarations in function scope.
1959 Patch from Oliver Fleischmann
1961 PP/10 Use sig_atomic_t for flags set from signal handlers.
1962 Check getgroups() return and improve debugging.
1963 Fixed developed for diagnosis in bug 927 (which turned out to be
1966 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1967 Patch from Mark Zealey.
1969 PP/12 Bugzilla 1056: Improved spamd server selection.
1970 Patch from Mark Zealey.
1972 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1973 Based on patch from Heiko Schlittermann.
1975 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1976 Patch from Uwe Doering, sign-off by Michael Haardt.
1978 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1979 variable declaration deep within a block. Bug and patch from
1982 PP/15 lookups-Makefile IRIX compatibility coercion.
1984 PP/16 Make DISABLE_DKIM build knob functional.
1986 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1987 Patch by Simon Arlott
1989 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1990 variable argument macros. Our copy now differs from upstream.
1996 TF/01 Failure to get a lock on a hints database can have serious
1997 consequences so log it to the panic log.
1999 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2000 controlled using the smtp_confirmation log selector.
2002 TF/03 Include the error message when we fail to unlink a spool file.
2004 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2005 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2006 for maintaining out-of-tree patches for some time.
2008 PP/01 Bugzilla 139: Documentation and portability issues.
2009 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2010 Handle per-OS dynamic-module compilation flags.
2012 PP/02 Let /dev/null have normal permissions.
2013 The 4.73 fixes were a little too stringent and complained about the
2014 permissions on /dev/null. Exempt it from some checks.
2015 Reported by Andreas M. Kirchwitz.
2017 PP/03 Report version information for many libraries, including
2018 Exim version information for dynamically loaded libraries. Created
2019 version.h, now support a version extension string for distributors
2020 who patch heavily. Dynamic module ABI change.
2022 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2023 privilege escalation vulnerability whereby the Exim run-time user
2024 can cause root to append content of the attacker's choosing to
2027 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2030 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2031 If dropping privileges for untrusted macros, we disabled normal logging
2032 on the basis that it would fail; for the Exim run-time user, this is not
2033 the case, and it resulted in successful deliveries going unlogged.
2034 Fixed. Reported by Andreas Metzler.
2040 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2041 only prepend for the Resent-* case. Fixes regression introduced in
2042 Exim 4.70 by NM/22 for Bugzilla 607.
2044 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2045 increasing numbers of administrators be bitten by this.
2047 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2049 PP/03 Bugzilla 994: added openssl_options main configuration option.
2051 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2053 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2055 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2057 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2058 without a peer certificate, leading to a segfault because of an
2059 assumption that peers always have certificates. Be a little more
2060 paranoid. Problem reported by Martin Tscholak.
2062 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2063 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2064 NB: ClamAV planning to remove STREAM in "middle of 2010".
2065 CL also introduces -bmalware, various -d+acl logging additions and
2066 more caution in buffer sizes.
2068 PP/09 Implemented reverse_ip expansion operator.
2070 PP/10 Bugzilla 937: provide a "debug" ACL control.
2072 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2074 PP/12 Bugzilla 973: Implement --version.
2076 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2078 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2080 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2082 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2083 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2084 ignore trailing whitespace.
2086 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2088 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2091 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2092 Notification from Dr Andrew Aitchison.
2094 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2095 ExtendedDetectionInfo response format.
2096 Notification from John Horne.
2098 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2101 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2102 XSL and documented dependency on system catalogs, with examples of how
2105 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2108 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2109 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2110 configuration file which is writeable by the Exim user or group.
2112 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2113 of configuration files to cover files specified with the -C option if
2114 they are going to be used with root privileges, not just the default
2117 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2118 option (effectively making it always true).
2120 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2121 files to be used while preserving root privileges.
2123 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2124 that rogue child processes cannot use them.
2126 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2127 run-time user, instead of root.
2129 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2130 Exim run-time user without dropping privileges.
2132 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2133 result string, instead of calling string_vformat() twice with the same
2136 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2137 for other users. Others should always drop root privileges if they use
2138 -C on the command line, even for a whitelisted configure file.
2140 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2142 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2148 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2149 $data_path, and $header_path variables; fixed documentation bugs and
2152 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2153 exipick to access non-standard spools, including the "frozen" queue
2156 NM/01 Bugzilla 965: Support mysql stored procedures.
2157 Patch from Alain Williams
2159 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2161 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2162 Patch from Andreas Metzler
2164 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2165 Patch from Kirill Miazine
2167 NM/05 Bugzilla 671: Added umask to procmail example.
2169 JJ/03 installed exipick 20100323.0, fixing doc bug
2171 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2172 directory. Notification and patch from Dan Rosenberg.
2174 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2176 TK/02 Improve log output when DKIM signing operation fails.
2178 MH/01 Treat the transport option dkim_domain as a colon separated
2179 list, not as a single string, and sign the message with each element,
2180 omitting multiple occurences of the same signer.
2182 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2183 Bugzilla 985, 986. Patch by Simon Arlott
2185 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2186 Patch by Simon Arlott
2188 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2189 MBX locking. Notification from Dan Rosenberg.
2195 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2197 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2199 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2201 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2203 NM/04 Bugzilla 915: Flag broken perl installation during build.
2209 TK/01 Added patch by Johannes Berg that expands the main option
2210 "spamd_address" if it starts with a dollar sign.
2212 TK/02 Write list of recipients to X-Envelope-Sender header when building
2213 the mbox-format spool file for content scanning (suggested by Jakob
2216 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2217 (http://www.dcc-servers.net/) support via dccifd. Activated by
2218 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2220 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2221 by Mark Daniel Reidel <mr@df.eu>.
2223 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2224 When building exim an external PCRE library is now needed -
2225 PCRE is a system library on the majority of modern systems.
2226 See entry on PCRE_LIBS in EDITME file.
2228 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2229 conversation. Added nologin parameter to request.
2230 Patch contributed by Kirill Miazine.
2232 TF/01 Do not log submission mode rewrites if they do not change the address.
2234 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2236 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2237 log files in place. Contributed by Roberto Lima.
2239 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2241 TF/03 Bugzilla 615: When checking the local_parts router precondition
2242 after a local_part_suffix or local_part_prefix option, Exim now
2243 does not use the address's named list lookup cache, since this
2244 contains cached lookups for the whole local part.
2246 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2247 Robert Millan. Documentation is in experimental-spec.txt.
2249 TF/04 Bugzilla 668: Fix parallel build (make -j).
2251 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2253 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2254 Patch provided by Jan Srzednicki.
2256 TF/05 Leading white space used to be stripped from $spam_report which
2257 wrecked the formatting. Now it is preserved.
2259 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2260 that they are available at delivery time.
2262 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2264 TF/08 TLS error reporting now respects the incoming_interface and
2265 incoming_port log selectors.
2267 TF/09 Produce a more useful error message if an SMTP transport's hosts
2268 setting expands to an empty string.
2270 NM/06 Bugzilla 744: EXPN did not work under TLS.
2271 Patch provided by Phil Pennock.
2273 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2274 Patch provided by Richard Godbee.
2276 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2277 acl_smtp_notquit, added index entry.
2279 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2280 Patch provided by Eugene Bujak.
2282 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2283 accept(). Patch provided by Maxim Dounin.
2285 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2286 Patch provided by Phil Pennock.
2288 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2290 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2291 Patch provided by Brad "anomie" Jorsch.
2293 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2294 Patch provided by Dean Brooks.
2296 TK/05 Add native DKIM support (does not depend on external libraries).
2298 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2299 Patch provided by Graeme Fowler.
2301 NM/16 Bugzilla 851: Documentation example syntax fix.
2303 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2305 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2308 NM/19 Bugzilla 745: TLS version reporting.
2309 Patch provided by Phil Pennock.
2311 NM/20 Bugzilla 167: bool: condition support.
2312 Patch provided by Phil Pennock.
2314 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2315 clients. Patch provided by Phil Pennock.
2317 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2318 Patch provided by Brad "anomie" Jorsch.
2320 NM/23 Bugzilla 687: Fix misparses in eximstats.
2321 Patch provided by Heiko Schlittermann.
2323 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2324 Patch provided by Heiko Schlittermann.
2326 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2327 plus update to original patch.
2329 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2331 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2332 Patch provided by David Brownlee.
2334 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2336 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2338 NM/30 Bugzilla 888: TLS documentation bugfixes.
2340 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2342 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2343 Unlike the original bugzilla I have changed all shell scripts in src tree.
2345 NM/33 Bugzilla 898: Transport filter timeout fix.
2346 Patch by Todd Rinaldo.
2348 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2349 Patch by Serge Demonchaux.
2351 NM/35 Bugzilla 39: Base64 decode bug fixes.
2352 Patch by Jakob Hirsch.
2354 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2356 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2358 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2360 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2366 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2367 ALT-N's libdkim that I have put here:
2368 http://duncanthrax.net/exim-experimental/
2370 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2371 were called 'true' and it seems that C99 defines that as a reserved
2372 keyword to be used with 'bool' variable types. That means you could
2373 not include C99-style headers which use bools without triggering
2374 build errors in sieve.c.
2376 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2377 as mailq or other aliases. Changed the --help handling significantly
2378 to do whats expected. exim_usage() emits usage/help information.
2380 SC/01 Added the -bylocaldomain option to eximstats.
2382 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2384 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2386 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2392 PH/01 Another patch from the Sieve maintainer.
2394 PH/02 When an IPv6 address is converted to a string for single-key lookup
2395 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2396 dots are used instead of colons so that keys in lsearch files need not
2397 contain colons. This was done some time before quoting was made available
2398 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2399 (notated using the quote facility) so as to distinguish them from IPv4
2400 keys. This meant that lookups for IP addresses in host lists did not work
2401 for iplsearch lookups.
2403 This has been fixed by arranging for IPv6 addresses to be expressed with
2404 colons if the lookup type is iplsearch. This is not incompatible, because
2405 previously such lookups could never work.
2407 The situation is now rather anomalous, since one *can* have colons in
2408 ordinary lsearch keys. However, making the change in all cases is
2409 incompatible and would probably break a number of configurations.
2411 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2414 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2415 conversion specification without a maximum field width, thereby enabling
2416 a rogue spamd server to cause a buffer overflow. While nobody in their
2417 right mind would setup Exim to query an untrusted spamd server, an
2418 attacker that gains access to a server running spamd could potentially
2419 exploit this vulnerability to run arbitrary code as the Exim user.
2421 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2422 $primary_hostname instead of what libspf2 thinks the hosts name is.
2424 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2425 a directory entry by the name of the lookup key. Previously, if a
2426 symlink pointed to a non-existing file or a file in a directory that
2427 Exim lacked permissions to read, a lookup for a key matching that
2428 symlink would fail. Now it is enough that a matching directory entry
2429 exists, symlink or not. (Bugzilla 503.)
2431 PH/03 The body_linecount and body_zerocount variables are now exported in the
2434 PH/04 Added the $dnslist_matched variable.
2436 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2437 This means they are set thereafter only if the connection becomes
2440 PH/06 Added the client_condition to authenticators so that some can be skipped
2441 by clients under certain conditions.
2443 PH/07 The error message for a badly-placed control=no_multiline_responses left
2444 "_responses" off the end of the name.
2446 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2448 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2449 (without spaces) instead of just copying the configuration text.
2451 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2453 PH/11 Added $max_received_linelength.
2455 PH/12 Added +ignore_defer and +include_defer to host lists.
2457 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2458 way in which PCRE > 7.0 is built.
2460 PH/14 Implemented queue_only_load_latch.
2462 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2463 MAIL command. The effect was to mangle the value on 64-bit systems.
2465 PH/16 Another patch from the Sieve maintainer.
2467 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2469 PH/18 If a system quota error occurred while trying to create the file for
2470 a maildir delivery, the message "Mailbox is full" was not appended to the
2471 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2472 only to a quota excession during the actual writing of the file.
2474 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2475 characters?) which causes problems in log lines. The DN values are now
2476 passed through string_printing() before being added to log lines.
2478 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2479 and InterBase are left for another time.)
2481 PH/21 Added message_body_newlines option.
2483 PH/22 Guard against possible overflow in moan_check_errorcopy().
2485 PH/23 POSIX allows open() to be a macro; guard against that.
2487 PH/24 If the recipient of an error message contained an @ in the local part
2488 (suitably quoted, of course), incorrect values were put in $domain and
2489 $local_part during the evaluation of errors_copy.
2495 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2496 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2499 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2500 issue a MAIL command.
2502 PH/02 In an ACL statement such as
2504 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2506 if a client was not listed at all, or was listed with a value other than
2507 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2508 the condition was not true (as it should be), so access was not denied.
2509 The bug was that the ! inversion was incorrectly passed on to the second
2510 item. This has been fixed.
2512 PH/03 Added additional dnslists conditions == and =& which are different from
2513 = and & when the dns lookup returns more than one IP address.
2515 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2516 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2518 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2519 FSYNC, which compiles an option called disable_fsync that allows for
2520 bypassing fsync(). The documentation is heavily laced with warnings.
2522 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2524 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2525 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2526 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2527 including adding "make clean"; (3) Added -fPIC when compiling the test
2528 dynamically loaded module, to get rid of a warning.
2530 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2531 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2532 = 0s. The bug is otherwise harmless.
2534 PH/07 There was a bug in the dovecot authenticator such that the value of
2535 $auth1 could be overwritten, and so not correctly preserved, after a
2536 successful authentication. This usually meant that the value preserved by
2537 the server_setid option was incorrect.
2539 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2541 PH/09 Installed PCRE release 7.0.
2543 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2544 run for batched SMTP input. It is now run at the start of every message
2545 in the batch. While fixing this I discovered that the process information
2546 (output by running exiwhat) was not always getting set for -bs and -bS
2547 input. This is fixed, and it now also says "batched" for BSMTP.
2549 PH/11 Added control=no_pipelining.
2551 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2552 patch, slightly modified), and move the expansion of helo_data till after
2553 the connection is made in the smtp transport (so it can use these
2556 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2558 PH/14 Added log_selector = +pid.
2560 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2562 PH/16 Add ${if forany and ${if forall.
2564 PH/17 Added dsn_from option to vary the From: line in DSNs.
2566 PH/18 Flush SMTP output before performing a callout, unless control =
2567 no_callout_flush is set.
2569 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2570 was true (the default) a successful delivery failed to delete the retry
2571 item, thus causing premature timeout of the address. The bug is now
2574 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2576 PH/21 Long custom messages for fakedefer and fakereject are now split up
2577 into multiline responses in the same way that messages for "deny" and
2578 other ACL rejections are.
2580 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2581 with slight modification.
2583 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2584 draft, changing the syntax and factoring some duplicate code".
2586 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2587 for deliveries of the second and subsequent messages over the same SMTP
2590 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2591 ${reduce, with only minor "tidies".
2593 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2595 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2596 expansion side effects.
2598 PH/27 When a message times out after an over-quota error from an Exim-imposed
2599 quota, the bounce message says "mailbox is full". This message was not
2600 being given when it was a system quota that was exceeded. It now should
2603 MH/03 Made $recipients available in local_scan(). local_scan() already has
2604 better access to the recipient list through recipients_list[], but
2605 $recipients can be useful in postmaster-provided expansion strings.
2607 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2608 in the case of a MAIL command with additional options following the
2609 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2610 were accidentally chopped off.
2612 PH/29 SMTP synchronization checks are implemented when a command is read -
2613 there is a check that no more input is waiting when there shouldn't be
2614 any. However, for some commands, a delay in an ACL can mean that it is
2615 some time before the response is written. In this time, more input might
2616 arrive, invalidly. So now there are extra checks after an ACL has run for
2617 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2618 pipelining has not been advertised.
2620 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2622 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2623 returned in the SMTP response when smtp_return_error_details was false.
2624 This has been fixed.
2626 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2627 instead of the C I/O that was originally supplied, because problems were
2628 reported on Solaris.
2630 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2631 Exim which did not show up earlier: it was assuming that a call to
2632 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2633 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2634 was a macro that became an assignment, so it seemed to work. This has
2635 changed to a proper function call with a void return, hence the compile
2636 error. Exim's code has been fixed.
2638 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2641 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2643 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2645 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2646 because it turns out to be too big or there is a timeout) there is no
2647 "Completed" line in the log. When some messages of this type were
2648 selected by exigrep, they were listed as "not completed". Others were
2649 picked up by some special patterns. I have improved the selection
2650 criteria to be more general.
2652 PH/38 The host_find_failed option in the manualroute router can now be set
2653 to "ignore", to completely ignore a host whose IP address cannot be
2654 found. If all hosts are ignored, the behaviour is controlled by the new
2655 host_all_ignored option.
2657 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2658 homing or because of multiple MX records with /mx) generated more than
2659 one IP address, and the following item turned out to be the local host,
2660 all the secondary addresses of the first item were incorrectly removed
2661 from the list, along with the local host and any following hosts (which
2662 is what is supposed to happen).
2664 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2665 whoever called Exim into the -H file. In the case of the daemon it was
2666 behaving confusingly. When first started, it used values for whoever
2667 started the daemon, but after a SIGHUP it used the Exim user (because it
2668 calls itself on a restart). I have changed the code so that it now always
2671 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2672 message are rejected with the same error (e.g. no authentication or bad
2673 sender address), and a DATA command is nevertheless sent (as can happen
2674 with PIPELINING or a stupid MUA), the error message that was given to the
2675 RCPT commands is included in the rejection of the DATA command. This is
2676 intended to be helpful for MUAs that show only the final error to their
2679 PH/42 Another patch from the Sieve maintainer.
2681 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2682 Eximstats - Fixed some broken HTML links and added missing column headers
2684 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2685 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2687 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2689 PH/43 Yet another patch from the Sieve maintainer.
2691 PH/44 I found a way to check for a TCP/IP connection going away before sending
2692 the response to the final '.' that terminates a message, but only in the
2693 case where the client has not sent further data following the '.'
2694 (unfortunately, this is allowed). However, in many cases there won't be
2695 any further data because there won't be any more messages to send. A call
2696 to select() can be used: if it shows that the input is "ready", there is
2697 either input waiting, or the socket has been closed. An attempt to read
2698 the next input character can distinguish the two cases. Previously, Exim
2699 would have sent an OK response which the client would never have see.
2700 This could lead to message repetition. This fix should cure that, at
2701 least in a lot of common cases.
2703 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2704 advertised in response to EHLO.
2710 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2711 fixed by 4.65/MH/01 (is this a record?) are fixed:
2713 (i) An empty string was always treated as zero by the numeric comparison
2714 operators. This behaviour has been restored.
2716 (ii) It is documented that the numeric comparison operators always treat
2717 their arguments as decimal numbers. This was broken in that numbers
2718 starting with 0 were being interpreted as octal.
2720 While fixing these problems I realized that there was another issue that
2721 hadn't been noticed. Values of message_size_limit (both the global option
2722 and the transport option) were treated as octal if they started with 0.
2723 The documentation was vague. These values are now always treated as
2724 decimal, and I will make that clear in the documentation.
2730 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2731 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2734 MH/01 Don't check that the operands of numeric comparison operators are
2735 integers when their expansion is in "skipping" mode (fixes bug
2736 introduced by 4.64-PH/07).
2738 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2739 child addresses, Exim now panics and dies. Previously, because the count
2740 is held in a short int, deliveries were likely to be lost. As such a
2741 large number of recipients for a single message is ridiculous
2742 (performance will be very, very poor), I have chosen to impose a limit
2743 rather than extend the field.
2749 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2750 leftover -K file (the existence of which was triggered by #402).
2751 While we were at it, introduced process PID as part of the -K
2752 filename. This should rule out race conditions when creating
2755 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2756 processing considerably. Previous code took too long for large mails,
2757 triggering a timeout which in turn triggers #401.
2759 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2760 in the DK code in transports.c. sendfile() is not really portable,
2761 hence the _LINUX specificness.
2763 TF/01 In the add_headers option to the mail command in an Exim filter,
2764 there was a bug that Exim would claim a syntax error in any
2765 header after the first one which had an odd number of characters
2768 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2769 callout verification, Exim cached a "reject" for the entire domain. This
2770 is correct for most verifications, but it is not correct for a recipient
2771 verification with use_sender or use_postmaster set, because in that case
2772 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2773 case of MAIL FROM:<> rejection from other early rejections (e.g.
2774 rejection of HELO). When verifying a recipient using a non-null MAIL
2775 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2776 Whatever the result of the callout, the value of the domain cache is
2777 left unchanged (for any other kind of callout, getting as far as trying
2778 RCPT means that the domain itself is ok).
2780 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2783 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2784 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2785 session, the daemon ignores ECONNECT errors and logs others; it now
2786 ignores EPIPE as well.
2788 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2789 (quoted-printable decoding).
2791 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2792 later the small subsequent patch to fix an introduced bug.
2794 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2796 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2798 PH/08 An error is now given if message_size_limit is specified negative.
2800 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2801 to be given (somewhat) arbitrary names.
2803 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2806 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2807 miscellaneous code fixes
2809 PH/10 Added the log_reject_target ACL modifier to specify where to log
2812 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2813 hostname. This is wrong, because it relates to the incoming message (and
2814 probably the interface on which it is arriving) and not to the outgoing
2815 callout (which could be using a different interface). This has been
2816 changed to use the value of the helo_data option from the smtp transport
2817 instead - this is what is used when a message is actually being sent. If
2818 there is no remote transport (possible with a router that sets up host
2819 addresses), $smtp_active_hostname is used.
2821 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2822 tweaks were necessary in order to get it to work (see also 21 below):
2823 (a) The code assumed that strncpy() returns a negative number on buffer
2824 overflow, which isn't the case. Replaced with Exim's string_format()
2826 (b) There were several signed/unsigned issues. I just did the minimum
2827 hacking in of casts. There is scope for a larger refactoring.
2828 (c) The code used strcasecmp() which is not a standard C function.
2829 Replaced with Exim's strcmpic() function.
2830 (d) The code set only $1; it now sets $auth1 as well.
2831 (e) A simple test gave the error "authentication client didn't specify
2832 service in request". It would seem that Dovecot has changed its
2833 interface. Fortunately there's a specification; I followed it and
2834 changed what the client sends and it appears to be working now.
2836 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2839 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2840 address A is aliased to B and C, where B exists and C does not. Without
2841 -v the output is "A verified" because verification stops after a
2842 successful redirection if more than one address is generated. However,
2843 with -v the child addresses are also verified. Exim was outputting "A
2844 failed to verify" and then showing the successful verification for C,
2845 with its parentage. It now outputs "B failed to verify", showing B's
2846 parentage before showing the successful verification of C.
2848 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2849 look up a TXT record in a specific list after matching in a combined
2852 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2853 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2854 they consult the DNS. I had assumed they would set it the way they
2855 wanted; and indeed my experiments on Linux seem to show that in some
2856 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2857 To be on the safe side, however, I have now made the interface to
2858 host_find_byname() similar to host_find_bydns(), with an argument
2859 containing the DNS resolver options. The host_find_byname() function now
2860 sets these options at its start, just as host_find_bydns() does. The smtp
2861 transport options dns_qualify_single and dns_search_parents are passed to
2862 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2863 of host_find_byname() use the default settings of RES_DEFNAMES
2864 (qualify_single) but not RES_DNSRCH (search_parents).
2866 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2867 spool_read_header() do less string testing, by means of a preliminary
2868 switch on the second character of optional "-foo" lines. (This is
2869 overdue, caused by the large number of possibilities that now exist.
2870 Originally there were few.) While I was there, I also converted the
2871 str(n)cmp tests so they don't re-test the leading "-" and the first
2872 character, in the hope this might squeeze out yet more improvement.
2874 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2875 flag allowing group syntax was set by the header_syntax check but not
2876 turned off, possible causing trouble later; (2) The flag was not being
2877 set at all for the header_verify test, causing "group"-style headers to
2878 be rejected. I have now set it in this case, and also caused header_
2879 verify to ignore an empty address taken from a group. While doing this, I
2880 came across some other cases where the code for allowing group syntax
2881 while scanning a header line wasn't quite right (mostly, not resetting
2882 the flag correctly in the right place). These bugs could have caused
2883 trouble for malformed header lines. I hope it is now all correct.
2885 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2886 with the "reply" argument non-NULL. The code, however (which originally
2887 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2888 but it didn't always do it. This confused somebody who was copying the
2889 code for some other use. I have removed all the tests.
2891 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2892 feature that was used to support insecure browsers during the U.S. crypto
2893 embargo. It requires special client support, and Exim is probably the
2894 only MTA that supported it -- and would never use it because real RSA is
2895 always available. This code has been removed, because it had the bad
2896 effect of slowing Exim down by computing (never used) parameters for the
2897 RSA_EXPORT functionality.
2899 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2900 authenticator to fail if there's a tab character in the incoming data
2901 (there should never be unless someone is messing about, as it's supposed
2902 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2903 if the connection is using TLS or if the remote IP is the same as the
2904 local IP, and the "valid-client-cert option" if a client certificate has
2907 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2908 authenticators. This can be used for authorization after authentication
2909 succeeds. (In the case of plaintext, it servers for both authentication
2912 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2913 if any retry times were supplied.
2915 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2916 connection, where there is no client IP address to check. In this
2917 situation, the verify now always succeeds.
2919 PH/25 Applied John Jetmore's -Mset patch.
2921 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2923 PH/27 In a string expansion for a processed (not raw) header when multiple
2924 headers of the same name were present, leading whitespace was being
2925 removed from all of them, but trailing whitespace was being removed only
2926 from the last one. Now trailing whitespace is removed from each header
2927 before concatenation. Completely empty headers in a concatenation (as
2928 before) are ignored.
2930 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2931 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2933 PH/29 [Removed. This was a change that I later backed out, and forgot to
2934 correct the ChangeLog entry (that I had efficiently created) before
2935 committing the later change.]
2937 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2938 address errors (4xx response to RCPT) over the same connection as other
2939 messages routed to the same hosts. Such deliveries are always "forced",
2940 so retry times are not inspected. This resulted in far too many retries
2941 for the affected addresses. The effect occurred only when there were more
2942 hosts than the hosts_max_try setting in the smtp transport when it had
2943 the 4xx errors. Those hosts that it had tried were not added to the list
2944 of hosts for which the message was waiting, so if all were tried, there
2945 was no problem. Two fixes have been applied:
2947 (i) If there are any address or message errors in an SMTP delivery, none
2948 of the hosts (tried or untried) are now added to the list of hosts
2949 for which the message is waiting, so the message should not be a
2950 candidate for sending over the same connection that was used for a
2951 successful delivery of some other message. This seems entirely
2952 reasonable: after all the message is NOT "waiting for some host".
2953 This is so "obvious" that I'm not sure why it wasn't done
2954 previously. Hope I haven't missed anything, but it can't do any
2955 harm, as the worst effect is to miss an optimization.
2957 (ii) If, despite (i), such a delivery is accidentally attempted, the
2958 routing retry time is respected, so at least it doesn't keep
2959 hammering the server.
2961 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2962 in ${readsocket because some servers need this prod.
2964 PH/32 Added some extra debug output when updating a wait-xxx database.
2966 PH/33 The hint "could be header name not terminated by colon", which has been
2967 given for certain expansion errors for a long time, was not being given
2968 for the ${if def:h_colon_omitted{... case.
2970 PH/34 The spec says: "With one important exception, whenever a domain list is
2971 being scanned, $domain contains the subject domain." There was at least
2972 one case where this was not true.
2974 PH/35 The error "getsockname() failed: connection reset by peer" was being
2975 written to the panic log as well as the main log, but it isn't really
2976 panic-worthy as it just means the connection died rather early on. I have
2977 removed the panic log writing for the ECONNRESET error when getsockname()
2980 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2981 runs only) independently of the message's sender address. This meant
2982 that, if the 4xx error was in fact related to the sender, a different
2983 message to the same recipient with a different sender could confuse
2984 things. In particular, this can happen when sending to a greylisting
2985 server, but other circumstances could also provoke similar problems.
2986 I have changed the default so that the retry time for these errors is now
2987 based a combination of the sender and recipient addresses. This change
2988 can be overridden by setting address_retry_include_sender=false in the
2991 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2992 remote server are returned as part of bounce messages. This was not
2993 happening for LMTP over a pipe (the lmtp transport), but now it is the
2994 same for both kinds of LMTP.
2996 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2997 in header lines that were in fact CNAMEs. This is no longer the case.
2999 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3000 and queue runs started by the daemon processed all messages. This has
3001 been fixed so that -R and -S can now usefully be given with -q<time>.
3003 PH/40 Import PCRE release 6.7 (fixes some bugs).
3005 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3007 PH/42 Give an error if -q is specified more than once.
3009 PH/43 Renamed the variables $interface_address and $interface_port as
3010 $received_ip_address and $received_port, to make it clear that these
3011 values apply to message reception, and not to the outgoing interface when
3012 a message is delivered. (The old names remain recognized, of course.)
3014 PH/44 There was no timeout on the connect() call when using a Unix domain
3015 socket in the ${readsocket expansion. There now is.
3017 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3018 be meaningful with "accept".
3020 SC/01 Eximstats V1.43
3021 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3023 SC/02 Eximstats V1.44
3024 Use a glob alias rather than an array ref in the generated
3025 parser. This improves both readability and performance.
3027 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3028 Collect SpamAssassin and rejection statistics.
3029 Don't display local sender or destination tables unless
3030 there is data to show.
3031 Added average volumes into the top table text output.
3033 SC/04 Eximstats V1.46
3034 Collect data on the number of addresses (recipients)
3035 as well as the number of messages.
3037 SC/05 Eximstats V1.47
3038 Added 'Message too big' to the list of mail rejection
3039 reasons (thanks to Marco Gaiarin).
3041 SC/06 Eximstats V1.48
3042 Mainlog lines which have GMT offsets and are too short to
3043 have a flag are now skipped.
3045 SC/07 Eximstats V1.49 (Alain Williams)
3046 Added the -emptyok flag.
3048 SC/08 Eximstats V1.50
3049 Fixes for obtaining the IP address from reject messages.
3051 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3052 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3053 whitespace changes from 4.64-PH/27
3055 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3058 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3059 are found, allow negative numbers in numeric criteria)
3061 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3063 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3064 to match changes made in 4.64-PH/43
3066 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3068 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3069 "domain" is actually an IP address, and force a failure. This locks out
3070 those revolvers/nameservers that support "A-for-A" lookups, in
3071 contravention of the specifications.
3073 PH/48 When a host name was looked up from an IP address, and the subsequent
3074 forward lookup of the name timed out, the host name was left in
3075 $sender_host_name, contrary to the specification.
3077 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3078 restricted to single-key lookups, Exim was not diagnosing an error if
3079 * or *@ was used with a query-style lookup.
3081 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3083 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3084 long ago, but noone interested enough thought of it. Let's just say that
3085 the "1.1" means that there are some new functions that weren't there at
3086 some point in the past.
3088 PH/51 Error processing for expansion failure of helo_data from an smtp
3089 transport during callout processing was broken.
3091 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3092 tested/used via the -bh/-bhc/-bs options.
3094 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3095 bug, fixed in subsequent PCRE releases).
3097 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3098 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3100 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3106 SC/01 Use a glob alias rather than an array ref in eximstats generated
3107 parser. This improves both readability and performance.
3109 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3110 Don't display local sender or destination tables in eximstats unless
3111 there is data to show.
3112 Added average volumes into the eximstats top table text output.
3114 SC/03 Collect data on the number of addresses (recipients) as well
3115 as the number of messages in eximstats.
3117 TF/01 Correct an error in the documentation for the redirect router. Exim
3118 does (usually) call initgroups() when daemonizing.
3120 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3121 with consistent privilege compared to when running as a daemon.
3123 TF/03 Note in the spec that $authenticated_id is not set for local
3124 submissions from trusted users.
3126 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3127 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3129 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3130 by adding some example configuration directives to the default
3131 configuration file. A little bit of work is required to uncomment the
3132 directives and define how usernames and passwords are checked, but
3133 there is now a framework to start from.
3135 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3136 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3137 without this. I don't know how relevant this is to other LDAP libraries.
3139 PH/02 Add the verb name to the "unknown ACL verb" error.
3141 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3143 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3145 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3146 directory not expanded when it should be if an expanded home directory
3147 was set for the address (which is overridden by the transport).
3149 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3152 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3153 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3154 because it is too late at that time, and has no effect.
3156 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3157 security issue with \' (bugzilla #107). I could not use the
3158 PQescapeStringConn() function, because it needs a PGconn value as one of
3161 PH/08 When testing addresses using -bt, indicate those final addresses that
3162 are duplicates that would not cause an additional delivery. At least one
3163 person was confused, thinking that -bt output corresponded to deliveries.
3164 (Suppressing duplicates isn't a good idea as you lose the information
3165 about possibly different redirections that led to the duplicates.)
3167 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3168 systems where poll() doesn't work, in particular OS X.
3170 PH/10 Added more information to debugging output for retry time not reached.
3172 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3173 operations in malware.c.
3175 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3178 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3179 syslog_duplication set false, log lines that would normally be written
3180 both the the main log and to the reject log were not written to syslog at
3183 PH/14 In the default configuration, change the use of "message" in ACL warn
3184 statements to "add_header".
3186 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3187 not followed by a command (e.g. "seen endif").
3189 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3190 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3193 PH/17 Added extra conditions to the default value of delay_warning_condition
3197 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3198 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3199 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3202 The Auto-Submitted: and various List- headers are standardised, whereas I
3203 don't think Precedence: ever was.
3205 PH/18 Refactored debugging code in route_finduser() to show more information,
3206 in particular, the error code if getpwnam() issues one.
3208 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3209 This is apparently needed in addition to the PH/07 change above to avoid
3210 any possible encoding problems.
3212 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3213 but not after initializing Perl.
3215 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3216 output them only if debugging. By default they are written stderr,
3217 apparently, which is not desirable.
3219 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3222 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3225 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3227 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3228 authenticated or an ident call has been made. Suppress the default
3229 values for $authenticated_id and $authenticated_sender (but permit -oMai
3230 and -oMas) when testing with -bh.
3232 PH/24 Re-jigged the order of the tests in the default configuration so that the
3233 tests for valid domains and recipients precede the DNS black list and CSA
3234 tests, on the grounds that those ones are more expensive.
3236 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3237 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3238 This bug exists in every version of Exim that I still have, right back to
3241 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3242 However, an attempt to turn on case-sensitivity in a regex key by
3243 including (?-i) didn't work because the subject string was already
3244 lowercased, and the effects were non-intuitive. It turns out that a
3245 one-line patch can be used to allow (?-i) to work as expected.
3251 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3252 other effects) broke the use of negated acl sub-conditions.
3254 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3257 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3258 "Deny" causes Exim to reject the incoming connection with a 554 error.
3259 Unfortunately, if there is a major crisis, such as a disk failure,
3260 tcp-wrappers gives "deny", whereas what one would like would be some
3261 kind of temporary error. A kludge has been added to help with this.
3262 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3263 554 error is used if errno is still zero or contains ENOENT (which occurs
3264 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3267 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3269 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3270 errors. Otherwise a message that provokes a temporary error (when other
3271 messages do not) can cause a whole host to time out.
3273 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3274 addresses were routed directly to files or pipes from a redirect router.
3275 File deliveries just didn't batch; pipe deliveries might have suffered
3278 PH/06 A failure to get a lock for a hints database would erroneously always say
3279 "Failed to get write lock", even when it was really a read lock.
3281 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3282 of 0600. This has been changed to use the value of the lockfile_mode
3283 option (which defaults to 0600).
3285 PH/08 Applied small patch from the Sieve maintainer.
3287 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3288 folder from quota calculations, a direct delivery into this folder messed
3289 up the contents of the maildirsize file. This was because the regex was
3290 used only to exclude .Trash (or whatever) when the size of the mailbox
3291 was calculated. There was no check that a delivery was happening into an
3292 excluded directory. This bug has been fixed by ignoring all quota
3293 processing for deliveries into excluded directories.
3295 PH/10 Added the maildirfolder_create_regex option to appendfile.
3301 PH/01 The code for finding all the local interface addresses on a FreeBSD
3302 system running IPv6 was broken. This may well have applied to all BSD
3303 systems, as well as to others that have similar system calls. The broken
3304 code found IPv4 interfaces correctly, but gave incorrect values for the
3305 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3306 that it would not match correctly against @[] and not recognize the IPv6
3309 PH/02 The ipliteral router was not recognizing addresses of the form user@
3310 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3312 PH/03 Added disable_ipv6.
3314 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3315 decoded form, because it is most often used to construct To: headers
3316 lines in autoreplies, and the decoded form may well be syntactically
3317 invalid. However, $reply_address has leading white space removed, and all
3318 newlines turned into spaces so that the autoreply transport does not
3321 PH/05 If group was specified without a user on a router, and no group or user
3322 was specified on a transport, the group from the router was ignored.
3324 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3325 for visible compile-time settings that can be used to change these
3326 numbers, for those that want even more. Backwards compatibility with old
3327 spool files has been maintained. However, going back to a previous Exim
3328 release will lost any variables that are in spool files.
3330 PH/07 Two small changes when running in the test harness: increase delay when
3331 passing a TCP/IP connection to a new process, in case the original
3332 process has to generate a bounce, and remove special handling of
3333 127.0.0.2 (sic), which is no longer necessary.
3335 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3336 be the same on different OS.
3338 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3341 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3342 whether --show-vars was specified or not
3344 JJ/02 exipick: Added support for new ACL variable spool format introduced
3347 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3348 syntactically invalid From: or Reply-to: line, and a filter used this to
3349 generate an autoreply, and therefore failed to obtain an address for the
3350 autoreply, Exim could try to deliver to a non-existent relative file
3351 name, causing unrelated and misleading errors. What now happens is that
3352 it logs this as a hard delivery error, but does not attempt to create a
3355 PH/11 The exinext utility has a -C option for testing purposes, but although
3356 the given file was scanned by exinext itself; it wasn't being passed on
3357 when Exim was called.
3359 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3360 an end-of-file indication when reading a command response.
3362 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3363 compiled. In many other places in Exim, IPv6 addresses are always
3364 recognized, so I have changed this. It also means that IPv4 domain
3365 literals of the form [IPV4:n.n.n.n] are now always recognized.
3367 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3368 used if the router is not running as root, for example, when verifying at
3369 ACL time, or when using -bh. The debugging output from this situation was
3370 non-existent - all you got was a failure to exec. I have made two
3373 (a) Failures to set uid/gid, the current directory, or a process leader
3374 in a subprocess such as that created by queryprogram now generate
3375 suitable debugging output when -d is set.
3377 (b) The queryprogram router detects when it is not running as root,
3378 outputs suitable debugging information if -d is set, and then runs
3379 the subprocess without attempting to change uid/gid.
3381 PH/15 Minor change to Makefile for building test_host (undocumented testing
3384 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3385 additional section of a DNS packet that returns MX or SRV records.
3386 Instead, it always explicitly searches for A/AAAA records. This avoids
3387 major problems that occur when a DNS server includes only records of one
3388 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3389 fixed another bug: if SRV records were looked up and the corresponding
3390 address records were *not* found in the additional section, the port
3391 values from the SRV records were lost.
3393 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3394 using the correct key (the original address) when searching the retry
3395 rules in order to find which one to use for generating the retry hint.
3397 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3398 adding the default one. Similarly, if it contains a Reply-To: header, the
3399 errors_reply_to option, if set, is not used.
3401 PH/19 When calculating a retry time, Exim used to measure the "time since
3402 failure" by looking at the "first failed" field in the retry record. Now
3403 it does not use this if it is later than than the arrival time of the
3404 message. Instead it uses the arrival time. This makes for better
3405 behaviour in cases where some deliveries succeed, thus re-setting the
3406 "first failed" field. An example is a quota failure for a huge message
3407 when small messages continue to be delivered. Without this change, the
3408 "time since failure" will always be short, possible causing more frequent
3409 delivery attempts for the huge message than are intended.
3410 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3412 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3413 $1, $2, $3) because the numerical variables can be reset during some
3414 expansion items (e.g. "match"), thereby losing the authentication data.
3416 PH/21 Make -bV show the size of off_t variables so that the test suite can
3417 decide whether to run tests for quotas > 2G.
3419 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3420 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3421 filecount value is greater than 2G or if a quota value is greater than 2G
3422 on a system where the size of off_t is not greater than 4, a panic error
3425 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3426 never match. The debug and -bh output now contains an explicit error
3427 message indicating a malformed IPv4 address or mask.
3429 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3430 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3431 PH/23 above applies.
3433 PH/25 Do not write to syslog when running in the test harness. The only
3434 occasion when this arises is a failure to open the main or panic logs
3435 (for which there is an explicit test).
3437 PH/26 Added the /no_tell option to "control=freeze".
3439 PH/27 If a host name lookup failed very early in a connection, for example, if
3440 the IP address matched host_lookup and the reverse lookup yielded a name
3441 that did not have a forward lookup, an error message of the form "no IP
3442 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3443 could be logged. Now it outputs the IP address instead of "NULL".
3445 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3446 allows the sender and the authenticated sender to be set when
3447 submitting a message from within Exim. Since child_open_exim() is
3448 documented for local_scan(), the new function should be too.
3450 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3451 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3452 results in an empty string is now treated as unset.
3454 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3456 PH/31 Added sender_verify_fail logging option.
3458 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3459 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3460 tidied the source and removed it altogether.
3462 PH/33 When a queue run was abandoned because the load average was too high, a
3463 log line was always written; now it is written only if the queue_run log
3464 selector is set. In addition, the log line for abandonment now contains
3465 information about the queue run such as the pid. This is always present
3466 in "start" and "stop" lines but was omitted from the "abandon" line.
3468 PH/34 Omit spaces between a header name and the colon in the error message that
3469 is given when verify = headers_syntax fails (if there are lots of them,
3470 the message gets confusing).
3472 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3473 names, as there are now some PTR records that contain slashes. This check
3474 is only to protect against broken name servers that fall over on strange
3475 characters, so the fact that it applies to all lookups doesn't matter.
3477 PH/36 Now that the new test suite is complete, we can remove some of the
3478 special code in Exim that was needed for the old test suite. For example,
3479 sorting DNS records because real resolvers return them in an arbitrary
3480 order. The new test suite's fake resolver always returns records in the
3483 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3484 bounces) except when queue_only is set, to avoid logging races between
3485 the different processes.
3487 PH/38 Panic-die if .include specifies a non-absolute path.
3489 PH/39 A tweak to the "H" retry rule from its user.
3491 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3492 a label. They prevented compilation on older perls.
3494 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3495 a warning to be raised on newish perls.
3497 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3498 on queue. Changes to match documented behaviour of showing count of
3499 messages matching specified criteria.
3501 PH/40 Changed the default ident timeout from 30s to 5s.
3503 PH/41 Added support for the use of login_cap features, on those BSD systems
3504 that have them, for controlling the resources used by pipe deliveries.
3506 PH/42 The content-scanning code uses fopen() to create files in which to put
3507 message data. Previously it was not paying any attention to the mode of
3508 the files. Exim runs with umask(0) because the rest of the code creates
3509 files with open(), and sets the required mode explicitly. Thus, these
3510 files were ending up world-writeable. This was not a big issue, because,
3511 being within the spool directory, they were not world-accessible. I have
3512 created a function called modefopen, which takes an additional mode
3513 argument. It sets umask(777), creates the file, chmods it to the required
3514 mode, then resets the umask. All the relevant calls to fopen() in the
3515 content scanning code have been changed to use this function.
3517 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3518 to 24 hours. This avoids potential overflow problems when processing G
3519 and H retry rules. I suspect nobody ever tinkers with this value.
3521 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3523 PH/45 When the plaintext authenticator is running as a client, the server's
3524 challenges are checked to ensure they are valid base64 strings. By
3525 default, the authentication attempt is cancelled if an invalid string is
3526 received. Setting client_ignore_invalid_base64 true ignores these errors.
3527 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3528 they are received. Thus, the responses can be made to depend on the
3529 challenges. If an invalid string is ignored, an empty string is placed in
3532 PH/46 Messages that are created by the autoreply transport now contains a
3533 References: header, in accordance with RFCs 2822 and 3834.
3535 PH/47 Added authenticated_sender_force to the smtp transport.
3537 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3539 PH/49 Installed latest patch from the Sieve maintainer.
3541 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3542 was also set, the appendfile transport was unnecessarily scanning a
3543 directory of message files (e.g. for maildir delivery) to find the count
3544 of files (along with the size), even though it did not need this
3545 information. It now does the scan only if it needs to find either the
3546 size of the count of files.
3548 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3550 PH/52 Two bugs concerned with error handling when the smtp transport is
3553 (i) Exim was not creating retry information for temporary errors given
3554 for individual recipients after the DATA command when the smtp transport
3555 was used in LMTP mode. This meant that they could be retried too
3556 frequently, and not timed out correctly.
3558 (ii) Exim was setting the flag that allows error details to be returned
3559 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3560 recipients that were returned after the DATA command.
3562 PH/53 This is related to PH/52, but is more general: for any failing address,
3563 when detailed error information was permitted to be returned to the
3564 sender, but the error was temporary, then after the final timeout, only
3565 "retry timeout exceeded" was returned. Now it returns the full error as
3566 well as "retry timeout exceeded".
3568 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3569 do this, and (what is worse) MTAs that accept it.
3571 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3572 will now be deprecated.
3574 PH/56 New os.c-cygwin from the Cygwin maintainer.
3576 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3577 formats (previously only available in exim formats via -bpr, -bpru,
3578 and -bpra. Now also available in native and exiqgrep formats)
3580 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3581 with very large, slow to parse queues
3583 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3585 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3587 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3588 responses to authentication challenges, though it was showing the
3589 challenges; (ii) I've removed the CR characters from the debug output for
3592 PH/58 Allow for the insertion of a newline as well as a space when a string
3593 is turned into more than one encoded-word during RFC 2047 encoding. The
3594 Sieve code now uses this.
3596 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3597 data_4xx, lost_connection, tls_required.
3599 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3600 message was being sent as an SMTP response.
3602 PH/61 Add -l and -k options to exicyclog.
3604 PH/62 When verifying, if an address was redirected to one new address, so that
3605 verification continued, and the new address failed or deferred after
3606 having set something in $address_data, the value of $address_data was not
3607 passed back to the ACL. This was different to the case when no
3608 redirection occurred. The value is now passed back in both cases.
3610 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3611 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3612 use login_cap.h, so on its own it isn't the distinguishing feature. The
3613 new name refers directly to the setclassresources() function.
3615 PH/65 Added configuration files for NetBSD3.
3617 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3619 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3620 is preferred over IPv4.
3622 PH/68 The bounce_return_message and bounce_return_body options were not being
3623 honoured for bounces generated during the reception of non-SMTP messages.
3624 In particular, this applied to messages rejected by the ACL. This bug has
3625 been fixed. However, if bounce_return_message is true and bounce_return_
3626 body is false, the headers that are returned for a non-SMTP message
3627 include only those that have been read before the error was detected.
3628 (In the case of an ACL rejection, they have all been read.)
3630 PH/69 The HTML version of the specification is now built in a directory called
3631 spec_html instead of spec.html, because the latter looks like a path with
3632 a MIME-type, and this confuses some software.
3634 PH/70 Catch two compiler warnings in sieve.c.
3636 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3637 function verify_get_ident() calls ip_connect() to connect a socket, but
3638 if the "connect()" function timed out, ip_connect() used to close the
3639 socket. However, verify_get_ident() also closes the socket later, and in
3640 between Exim writes to the log, which may get opened at this point. When
3641 the socket was closed in ip_connect(), the log could get the same file
3642 descriptor number as the socket. This naturally causes chaos. The fix is
3643 not to close the socket in ip_connect(); the socket should be closed by
3644 the function that creates it. There was only one place in the code where
3645 this was missing, in the iplookup router, which I don't think anybody now
3646 uses, but I've fixed it anyway.
3648 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3649 well as to direct DNS lookups. Otherwise the handling of names in host
3650 lists is inconsistent and therefore confusing.
3656 PH/01 Two changes to the default runtime configuration:
3658 (1) Move the checks for relay_from_hosts and authenticated clients from
3659 after to before the (commented out) DNS black list checks.
3661 (2) Add control=submission to the relay_from_hosts and authenticated
3662 clients checks, on the grounds that messages accepted by these
3663 statements are most likely to be submissions.
3665 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3667 (1) Generate an error if the third argument for the ${prvs expansion is
3670 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3673 (3) Reset the variables that are obtained from the first argument of
3674 ${prvscheck and used in the second argument before leaving the code,
3675 because their memory is reclaimed, so using them afterwards may do
3678 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3679 one (it's much easier than Tom thought :-).
3681 (5) Because of (4), we can now allow for the use of $prvscheck_result
3682 inside the third argument.
3684 PH/03 For some reason, the default setting of PATH when running a command from
3685 a pipe transport was just "/usr/bin". I have changed it to
3688 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3689 anything to be listed in the output from -bV.
3691 PH/05 When a filter generated an autoreply, the entire To: header line was
3692 quoted in the delivery log line, like this:
3694 => >A.N.Other <ano@some.domain> <original@ddress> ...
3696 This has been changed so that it extracts the operative address. There
3697 may be more than one such address. If so, they are comma-separated, like
3700 => >ano@some.domain,ona@other.domain <original@ddress> ...
3702 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3703 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3704 not being looked up in the rDNS to get a host name, Exim was showing the
3705 IP address twice in Received: lines, even though the IP addresses were
3706 identical. For example:
3708 Received: from [1.2.3.4] (helo=[1.2.3.4])
3710 However, if the real host name was known, it was omitting the HELO data
3711 if it matched the actual IP address. This has been tidied up so that it
3712 doesn't show the same IP address twice.
3714 PH/07 When both +timestamp and +memory debugging was on, the value given by
3715 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3716 called by the debug printing, thereby overwriting the timestamp buffer.
3717 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3719 PH/08 When the original message was included in an autoreply transport, it
3720 always said "this is a copy of the message, including all the headers",
3721 even if body_only or headers_only was set. It now gives an appropriate
3724 PH/09 Applied a patch from the Sieve maintainer which:
3726 o fixes some comments
3727 o adds the (disabled) notify extension core
3728 o adds some debug output for the result of if/elsif tests
3729 o points to the current vacation draft in the documentation
3730 and documents the missing references header update
3734 o fixes a bug in processing the envelope test (when testing
3735 multiple envelope elements, the last element determined the
3738 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3739 Electronic Mail") by including:
3741 Auto-submitted: auto-generated
3743 in the messages that it generates (bounce messages and others, such as
3744 warnings). In the case of bounce messages for non-SMTP messages, there was
3745 also a typo: it was using "Auto_submitted" (underscore instead of
3746 hyphen). Since every message generated by Exim is necessarily in response
3747 to another message, thes have all been changed to:
3749 Auto-Submitted: auto-replied
3751 in accordance with these statements in the RFC:
3753 The auto-replied keyword:
3755 - SHOULD be used on messages sent in direct response to another
3756 message by an automatic process,
3758 - MUST NOT be used on manually-generated messages,
3760 - MAY be used on Delivery Status Notifications (DSNs) and Message
3761 Disposition Notifications (MDNs),
3763 - MUST NOT be used on messages generated by automatic or periodic
3764 processes, except for messages which are automatic responses to
3767 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3768 to the default Received: header definition.
3770 PH/12 Added log selector acl_warn_skipped (default on).
3772 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3773 variables because (a) they are in the wrong storage pool and (b) even if
3774 they were copied, it wouldn't work properly because of the caching.
3776 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3777 checking when decoding. Apparently there are clients that generate
3778 overlong encoded strings. Why am I not surprised?
3780 PH/15 If the first argument of "${if match_address" was not empty, but did not
3781 contain an "@" character, Exim crashed. Now it writes a panic log message
3782 and treats the condition as false.
3784 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3786 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3787 extension "envelope-auth". The code is finished and in agreement with
3788 other implementations, but there is no documentation so far and in fact,
3789 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3790 not changing the active code.
3792 Print executed "if" and "elsif" statements when debugging is used. This
3793 helps a great deal to understand what a filter does.
3795 Document more things not specified clearly in RFC3028. I had all this
3796 sorted out, when out of a sudden new issues came to my mind. Oops."
3798 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3801 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3802 canonical form (as documented). However, after a host name lookup from
3803 the IP address, check_host() was doing a simple string comparison with
3804 addresses acquired from the DNS when checking that the found name did
3805 have the original IP as one of its addresses. Since any found IPv6
3806 addresses are likely to be in abbreviated form, the comparison could
3807 fail. Luckily, there already exists a function for doing the comparison
3808 by converting both addresses to binary, so now that is used instead of
3809 the text comparison.
3811 PH/20 There was another similar case to PH/19, when a complete host name was
3812 given in a host list; looking up its IP address could give an abbreviated
3813 form, whereas the current host's name might or might not be abbreviated.
3814 The same fix has been applied.
3820 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3821 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3824 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3825 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3827 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3829 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3830 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3831 Also, for "Auto-Submitted", treat anything other than "no" as
3832 non-personal, in accordance with RFC 3834. (Previously it treated
3833 anything starting "auto-" as non-personal.)
3835 TF/01 The control=submission/name=... option had a problem with syntax
3836 errors if the name included a slash character. The /name= option
3837 now slurps the rest of the string, so it can include any characters
3838 but it must come last in the list of options (after /sender_retain
3841 PH/05 Some modifications to the interface to the fake nameserver for the new
3849 TK/01 Added the "success_on_redirect" address verification option. See
3850 NewStuff for rationale and an example.
3852 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3854 PH/02 Patch to exigrep to allow it to work on syslog lines.
3856 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3857 fread() to skip over the body file's header line, because in Cygwin the
3858 header line is locked and is inaccessible.
3860 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3861 co-exist for some time) to make it clear that it is the Exim ID that is
3862 referenced, not the Message-ID: header line.
3864 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3865 string_format() function, because snprintf() does not exist on all
3868 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3869 ${stat: expansion item.
3871 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3872 protocol synchronization error", to keep the pedants happy.
3874 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3875 well as for IRIX systems, when gcc is being used. See the host.c source
3878 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3880 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3883 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3884 transport and to the smtp transport in LMTP mode.
3886 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3888 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3889 run a filter in a subprocess. This could lead to confusion in subsequent
3890 lookups in the parent process. There should also be a search_tidyup() at
3891 the end of the subprocess.
3893 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3894 only if the host matched helo_try_verify_hosts, which caused the
3895 verification to occur when the EHLO/HELO command was issued. The ACL just
3896 tested the remembered result. Now, if a previous verification attempt has
3897 not happened, "verify = helo" does it there and then.
3899 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3901 TK/03 Fix log output including CR from clamd.
3903 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3904 exist provoked a memory error which could cause a segfault.
3906 PH/15 Installed PCRE 6.2
3908 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3910 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3911 of the problem. Specifically, suggested +O2 rather than +O1 for the
3914 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3916 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3917 record, the port was not correctly logged when the outgoing_port log
3918 selector was set (it logged the transort's default port).
3920 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3921 fallback_hosts, and "hosts" in the smtp transport.
3923 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3924 host errors such as "Connection refused".
3926 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3927 authentication with radiusclient 0.4.9:
3929 - Error returned from rc_read_config was caught wrongly
3930 - Username/password not passed on to radius server due to wrong length.
3932 The presumption is that some radiusclient API changes for 4.51/PH/17
3933 were not taken care of correctly. The code is still untested by me (my
3934 Linux distribution still has 0.3.2 of radiusclient), but it was
3935 contributed by a Radius user.
3937 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3938 expanding the "port" option of the smtp transport.
3940 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3941 while reading a MIME header. Thanks to Tom Hughes for a patch.
3943 PH/24 Include config.h inside local_scan.h so that configuration settings are
3946 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3947 that in an ACL for RCPT (for example), you can examine exactly what was
3950 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3951 commands, but it was not correctly comparing the address with the actual
3952 client host address. Thus, it would show the EHLO address in Received:
3953 header lines when this was not necessary.
3955 PH/27 Added the % operator to ${eval:}.
3957 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3958 it should be ignoring failures (because with -C, for example, it has lost
3959 privilege). It wasn't ignoring creation failures other than "already
3962 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3965 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3966 because an input line was too long, either on its own, or by virtue of
3967 too many continuations, the temporary file was not being removed, and the
3968 return code was incorrect.
3970 PH/31 Missing "BOOL" in function definition in filtertest.c.
3972 PH/32 Applied Sieve patches from the maintainer.
3974 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3976 PH/33 Added "verify = not_blind".
3978 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3979 Local/Makefile (with some defaults set). These are used in built scripts
3980 such as exicyclog, but they have never been used in the exim_install
3981 script (though there are many overriding facilities there). I have
3982 arranged that the exim_install script now takes note of these two
3985 PH/35 Installed configuration files for Dragonfly.
3987 PH/36 When a locally submitted message by a trusted user did not contain a
3988 From: header, and the sender address was obtained from -f or from an SMTP
3989 MAIL command, and the trusted user did not use -F to supply a sender
3990 name, $originator_name was incorrectly used when constructing a From:
3991 header. Furthermore, $originator_name was used for submission mode
3992 messages from external hosts without From: headers in a similar way,
3993 which is clearly wrong.
3995 PH/37 Added control=suppress_local_fixups.
3997 PH/38 When log_selector = +received_sender was set, and the addition of the
3998 sender made the log line's construction buffer exactly full, or one byte
3999 less than full, an overflow happened when the terminating "\n" was
4002 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4003 when the result of a list match is failure because a DNS lookup failed.
4005 PH/40 RM_COMMAND is now used in the building process.
4007 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4008 the "build-* directories that it finds.
4010 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4011 address in a domain literal was a prefix of an interface address.
4013 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4014 when verifying a sender address, unless rewrite_headers is false.
4016 PH/44 Wrote a long comment about why errors_to addresses are verified as
4017 recipients, not senders.
4019 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4020 the ratelimit ACL was added.
4022 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4024 PH/46 Added extra information about PostgreSQL errors to the error string.
4026 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4027 avoiding the need to install special zones in a real server. This is
4028 backwards compatible; if it can't find the fake resolver, it drops back.
4029 Thus, both old and new test suites can be run.
4031 TF/02 Added util/ratelimit.pl
4033 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4034 clock is set back in time.
4036 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4037 Candler <B.Candler@pobox.com>.
4039 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4040 OK for addresses that are the result of redirections.
4042 PH/48 A number of further additions for the benefit of the new test suite,
4043 including a fake gethostbyname() that interfaces to the fake DNS resolver
4046 TF/06 The fix for widen_domains has also been applied to qualify_single and
4047 search_parents which are the other dnslookup options that can cause
4050 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4053 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4055 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4056 service" if it encountered an MX record with an empty target hostname.
4057 The message is now "an MX or SRV record indicated no SMTP service".
4059 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4060 if the DNS of the sending site is misconfigured. This is quite a
4061 common situation. This change restores the behaviour of treating a
4062 helo verification defer as a failure.
4064 PH/51 If self=fail was set on a router, the bounce message did not include the
4065 actual error message.
4071 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4073 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4074 command itself ended in error, the underlying message about the transport
4075 filter timeout was being overwritten with the pipe command error. Now the
4076 underlying error message should be appended to the second error message.
4078 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4080 PH/02 Reduce the amount of output that "make" produces by default. Full output
4081 can still be requested.
4083 PH/03 The warning log line about a condition test deferring for a "warn" verb
4084 was being output only once per connection, rather than after each
4085 occurrence (because it was using the same function as for successful
4086 "warn" verbs). This seems wrong, so I have changed it.
4088 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4089 it should not have, which might have caused a crash in the right
4090 circumstances, but probably never did.
4092 PH/04 Installed a modified version of Tony Finch's patch to make submission
4093 mode fix the return path as well as the Sender: header line, and to
4094 add a /name= option so that you can make the user's friendly name appear
4097 TF/03 Added the control = fakedefer ACL modifier.
4099 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4100 Mark Lowes for thorough testing.
4102 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4104 TK/03 Merged latest SRS patch from Miles Wilton.
4106 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4107 with the definition in sysexits.h (which is #included earlier).
4108 Fortunately, Exim does not actually use EX_OK. The code used to try to
4109 preserve the sysexits.h value, by assuming that macro definitions were
4110 scanned for macro replacements. I have been disabused of this notion,
4111 so now the code just undefines EX_OK before #including unistd.h.
4113 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4114 in the smtp transport. When a block could not be written in a single
4115 write() function, the timeout was being re-applied to each part-write.
4116 This seems wrong - if the receiver was accepting one byte at a time it
4117 would take for ever. The timeout is now adjusted when this happens. It
4118 doesn't have to be particularly precise.
4120 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4121 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4123 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4124 without a domain if the check to <postmaster@domain> fails.
4126 SC/01 Eximstats: added -xls and the ability to specify output files
4127 (patch written by Frank Heydlauf).
4129 SC/02 Eximstats: use FileHandles for outputting results.
4131 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4133 SC/04 Eximstats: fixed display of large numbers with -nvr option
4135 SC/05 Eximstats: fixed merging of reports with empty tables.
4137 SC/06 Eximstats: added the -include_original_destination flag
4139 SC/07 Eximstats: removed tabs and trailing whitespace.
4141 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4143 TK/06 MBOX spool code: Add real "From " MBOX separator line
4144 so the .eml file is really in mbox format (even though
4145 most programs do not really care). Patch from Alex Miller.
4147 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4148 The latter is generated from $received_to and is only set if the
4149 message has one envelope recipient. SA can use these headers,
4150 obviously out-of-the-box. Patch from Alex Miller.
4152 PH/08 The ${def test on a variable was returning false if the variable's
4153 value was "0", contrary to what the specification has always said!
4154 The result should be true unless the variable is empty.
4156 PH/09 The syntax error of a character other than { following "${if
4157 def:variable_name" (after optional whitespace) was not being diagnosed.
4158 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4159 accidental colon was present, for example, could give incorrect results.
4161 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4162 result is used (not including appendfile, where other changes are about
4165 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4166 This involved changing a lot of size variables from int to off_t. It
4167 should work with maildirs and everything.
4169 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4170 spamd dying while we are connected to it.
4172 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4175 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4176 function for BDB 4.3.
4178 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4180 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4181 the sources for PCRE have been re-arranged and more files are now
4184 PH/15 The code I had for printing potentially long long variables in PH/11
4185 above was not the best (it lost precision). The length of off_t variables
4186 is now inspected at build time, and an appropriate printing format (%ld
4187 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4188 to be "long long int" or "long int". This is needed for the internal
4189 formatting function string_vformat().
4191 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4192 the configuration file to be ":syslog", then the script "guesses" where
4193 the logs files are, rather than using the compiled in default. In our
4194 case the guess is not the same as the compiled default, so the script
4195 suddenly stopped working when I started to use syslog. The patch checks
4196 to see if log_file_path is "". If so, it attempts to read it from exim
4197 with no configuration file to get the compiled in version, before it
4198 falls back to the previous guessing code."
4200 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4201 implementing BATV in an Exim configuration. See NewStuff for the gory
4204 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4205 Makefile that are specific to HP-UX.
4207 PH/18 If the "use_postmaster" option was set for a recipient callout together
4208 with the "random" option, the postmaster address was used as the MAIL
4209 FROM address for the random test, but not for the subsequent recipient
4210 test. It is now used for both.
4212 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4213 patch removes a few documentation additions to RFC 3028, because the
4214 latest draft now contains them. It adds the new en;ascii-case comparator
4215 and a new error check for 8bit text in MIME parts. Comparator and
4216 require names are now matched exactly. I enabled the subaddress
4217 extension, but it is not well tested yet (read: it works for me)."
4219 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4220 rework some of the code of TK/09 above to avoid the hardwired use of
4221 "%lld" and "long long". Replaced the call to snprintf() with a call to
4224 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4225 records point to non-existent hosts", "retry timeout exceeded", and
4226 "retry time not reached for any host after a long failure period".
4228 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4229 experimental DomainKeys support:
4231 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4232 (2) On an error such as an illegally used "control", the wrong name for
4233 the control was given.
4235 These problems did NOT occur unless DomainKeys support was compiled.
4237 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4239 PH/24 Added ${if match_ip condition.
4241 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4242 databases so that it will be absolutely obvious if a crash occurs in the
4243 DB library. This is a regular occurrence (often caused by mis-matched
4246 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4247 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4248 picked up on a user's system that detects such things. There doesn't seem
4249 to be a gcc warning option for this - only an attribute that has to be
4250 put on the function's prototype. It seems that in Fedora Core 4 they have
4251 set this on a number of new functions. No doubt there will be more in due
4254 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4255 specify a transport. However, if an address that was verified by such a
4256 router was the subject of a callout, Exim crashed because it tried to
4257 read the rcpt_include_affixes from the non-existent transport. Now it
4258 just assumes that the setting of that option is false. This bug was
4259 introduced by 4.51/PH/31.
4261 PH/28 Changed -d+all to exclude +memory, because that information is very
4262 rarely of interest, but it makes the output a lot bigger. People tend to
4263 do -d+all out of habit.
4265 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4266 code in os-type was giving problems when libc.so lives in lib64, like on
4269 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4270 aren't the modern standard, and it seems that some systems' include files
4271 don't always have them. Exim was already checking for some of the newer
4272 ones like T_AAAA, and defining it itself. I've added checks for all the
4273 record types that Exim uses.
4275 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4276 not automatically generating a new one, as it is supposed to. This
4277 prevented TLS from working. If the file did exist, but contained invalid
4278 data, a new version was generated, as expected. It was only the case of a
4279 non-existent file that was broken.
4281 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4282 with a change in libdomainkeys > 0.64.
4284 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4285 from DNS. If the selector record carries the flag, it now has
4286 precedence over the domain-wide flag.
4288 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4290 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4291 the use of a port name that isn't defined in /etc/services) occurred, the
4292 message was deferred as in a normal delivery, and thus remained on the
4293 spool, instead of being failed because of the mua_wrapper setting. This
4294 is now fixed, and I tidied up some of the mua_wrapper messages at the
4297 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4298 the messages in a hash of arrays rather than using individual hashes.
4299 This is a bit cleaner and results in dramatic memory savings, albeit
4300 at a slight CPU cost.
4302 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4303 as requested by Marc Sherman.
4305 SC/10 Eximstats: added histograms for user specified patterns as requested
4308 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4310 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4311 fopen() in the content-scanning modules that did not already have it.
4317 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4318 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4320 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4322 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4324 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4325 to test Sieve filters that use "vacation".
4327 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4328 that changes the way the GnuTLS parameters are stored in the cache file.
4329 The new format can be generated externally. For backward compatibility,
4330 if the data in the cache doesn't make sense, Exim assumes it has read an
4331 old-format file, and it generates new data and writes a new file. This
4332 means that you can't go back to an older release without removing the
4335 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4336 work if there are any delivery delays because "one_time" forces the
4337 parent to be marked "delivered", so its unseen clone is never tried
4338 again. For this reason, Exim now forbids the simultaneous setting of
4341 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4342 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4343 entry if you want to know the details. The fix, however, neglected to
4344 consider the case when local delivery batching is involved. The test for
4345 "previously delivered" was not happening when checking to see if an
4346 address could be batched with a previous (undelivered) one; under
4347 certain circumstances this could lead to multiple deliveries to the same
4350 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4351 in its include files, and this causes problems building Exim.
4353 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4354 header_syntax) but Exim was just ignoring anything given after a slash.
4355 In particular, this caused confusion with an attempt to use "verify =
4356 reverse_host_lookup/defer_ok". An error is now given when options are
4357 supplied for verify items that do not have them. (Maybe reverse_host_
4358 lookup should have a defer_ok option, but that's a different point.)
4360 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4361 defined by RFC 821) to 2048, because there were problems with some AUTH
4362 commands, and RFC 1869 says the size should be increased for extended
4363 SMTP commands that take arguments.
4365 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4368 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4369 "unknown" error; now it says that the functionality isn't in the binary.
4371 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4372 an address' error message when a string expansion fails (syntax or
4373 whatever). Otherwise the password may appear in the log. Following change
4374 PH/42 below, there is no longer a chance of it appearing in a bounce
4377 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4379 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4380 of its addresses was ever tried. (Bugzilla bug #2.)
4382 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4383 the result incorrectly in the debug output. (It correctly added a newline
4384 to what was transported.)
4386 TF/01 Added $received_time.
4388 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4389 commented out examples of how to interface to a virus scanner and to
4390 SpamAssassin. Also added commented examples of av_scanner and
4391 spamd_address settings.
4393 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4394 and controls are allowed in which ACLs. There were a couple of minor
4395 errors. Some of the entries in the conditions table (which is a table of
4396 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4397 negation of where the condition IS allowed.
4399 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4401 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4402 header file does not have a version number, so I've had to invent a new
4403 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4404 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4405 radiusclient), but it was contributed by a Radius user.
4407 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4408 files or directories, for OpenSSL.
4410 PH/19 When an Exim process that is running as root has to create an Exim log
4411 file, it does so in a subprocess that runs as exim:exim so as to get the
4412 ownership right at creation (otherwise, other Exim processes might see
4413 the file with the wrong ownership). There was no test for failure of this
4414 fork() call, which would lead to the process getting stuck as it waited
4415 for a non-existent subprocess. Forks do occasionally fail when resources
4416 run out. I reviewed all the other calls to fork(); they all seem to check
4419 PH/20 When checking for unexpected SMTP input at connect time (before writing
4420 the banner), Exim was not dealing correctly with a non-positive return
4421 from the read() function. If the client had disconnected by this time,
4422 the result was a log entry for a synchronization error with an empty
4423 string after "input=" when read() returned zero. If read() returned -1
4424 (an event I could not check), uninitialized data bytes were printed.
4425 There were reports of junk text (parts of files, etc) appearing after
4428 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4430 PH/22 Added support for macro redefinition, and (re)definition in between
4431 driver and ACL definitions.
4433 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4434 forgetting to use the resulting value; it was using the unexpanded value.
4436 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4437 hadn't been configured. The fix is from Juergen Kreileder, who
4438 understands it better than I do:
4440 "Here's what I see happening with three configured cyrus_sasl
4441 authenticators configured (plain, login, cram-md5):
4443 On startup auth_cyrus_sasl_init() gets called for each of these.
4444 This means three calls to sasl_listmech() without a specified mech_list.
4445 => SASL tests which mechs of all available mechs actually work
4446 => three warnings about OTP not working
4447 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4449 With the patch, sasl_listmech() also gets called three times. But now
4450 SASL's mech_list option is set to the server_mech specified in the the
4451 authenticator. Or in other words, the answer from sasl_listmech()
4452 gets limited to just the mech you're testing for (which is different
4454 => the return list contains just 'plain' or 'login', 'cram-md5' or
4455 nothing depending on the value of ob->server_mech.
4457 I've just tested the patch: Authentication still works fine,
4458 unavailable mechs specified in the exim configuration are still
4459 caught, and the auth.log warnings about OTP are gone."
4461 PH/25 When debugging is enabled, the contents of the command line are added
4462 to the debugging output, even when log_selector=+arguments is not
4465 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4466 answer is "GNU", and only if the return is "GNU/something" is the answer
4469 PH/27 $acl_verify_message is now set immediately after the failure of a
4470 verification in an ACL, and so is available in subsequent modifiers. In
4471 particular, the message can be preserved by coding like this:
4473 warn !verify = sender
4474 set acl_m0 = $acl_verify_message
4476 Previously, $acl_verify_message was set only while expanding "message"
4477 and "log_message" when a very denied access.
4479 PH/28 Modified OS/os.c-Linux with
4481 -#ifndef OS_LOAD_AVERAGE
4482 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4484 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4485 nomenclature these days.)
4487 PH/29 Installed patch from the Sieve maintainer that adds the options
4488 sieve_useraddress and sieve_subaddress to the redirect router.
4490 PH/30 In these circumstances:
4491 . Two addresses routed to the same list of hosts;
4492 . First host does not offer TLS;
4493 . First host accepts first address;
4494 . First host gives temporary error to second address;
4495 . Second host offers TLS and a TLS session is established;
4496 . Second host accepts second address.
4497 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4498 and peerdn, if requested) that were in fact used only for the second
4501 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4502 attention to any local part prefix or suffix that was matched by the
4503 router that accepted the address. It now behaves in the same way as it
4504 does for delivery: the affixes are removed from the local part unless
4505 rcpt_include_affixes is set on the transport.
4507 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4508 timeout during the DATA phase of an incoming message.
4510 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4511 applied a patch sanctioned by the Sieve maintainer.
4513 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4514 the uid or gid is negative. A case of a negative gid caused this to be
4515 noticed. The fix allows for either to be negative.
4517 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4518 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4521 PH/36 The change PH/12 above was broken. Fixed it.
4523 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4524 the grounds that routing the same address twice would always produce the
4525 same answer. This might have been true once, but it is certainly no
4526 longer true now. Routing a child address may depend on the previous
4527 routing that produced that child. Some complicated redirection strategies
4528 went wrong when messages had multiple recipients, and made Exim's
4529 behaviour dependent on the order in which the addresses were given.
4531 I have moved the duplicate checking until after the routing is complete.
4532 Exim scans the addresses that are assigned to local and remote
4533 transports, and removes any duplicates. This means that more work will be
4534 done, as duplicates will always all be routed, but duplicates are
4535 presumably rare, so I don't expect this is of any significance.
4537 For deliveries to pipes, files, and autoreplies, the duplicate checking
4538 still happens during the routing process, since they are not going to be
4541 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4542 It corrects a timeout issue with spamd. This is Ian's comment: "The
4543 background is that sometimes spamd either never reads data from a
4544 connection it has accepted, or it never writes response data. The exiscan
4545 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4546 blindly assumes that writes won't block so it may never time out."
4548 PH/39 Allow G after quota size as well as K and M.
4550 PH/40 The value set for $authenticated_id in an authenticator may not contain
4551 binary zeroes or newlines because the value is written to log lines and
4552 to spool files. There was no check on this. Now the value is run through
4553 the string_printing() function so that such characters are converted to
4554 printable escape sequences.
4556 PH/41 $message_linecount is a new variable that contains the total number of
4557 lines in the message. Compare $body_linecount, which is the count for the
4560 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4561 bounce and delay warning messages, except in certain special cases, which
4564 (a) An SMTP error message from a remote host;
4565 (b) A message specified in a :fail: redirection;
4566 (c) A message specified in a "fail" command in a system filter;
4567 (d) A message specified in a FAIL return from the queryprogram router;
4568 (e) A message specified by the cannot_route_message router option.
4570 In these cases only, Exim does include the error details in bounce and
4571 warning messages. There are also a few cases where bland messages such
4572 as "unrouteable address" or "local delivery error" are given.
4574 PH/43 $value is now also set for the "else" part of a ${run expansion.
4576 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4577 being worked on, but at least Exim now implements the latest version to
4580 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4581 process to complete was treated as a delivery failure, a timeout while
4582 writing the message to the pipe was logged, but erroneously treated as a
4583 successful delivery. Such timeouts include transport filter timeouts. For
4584 consistency with the overall process timeout, these timeouts are now
4585 treated as errors, giving rise to delivery failures by default. However,
4586 there is now a new Boolean option for the pipe transport called
4587 timeout_defer, which, if set TRUE, converts the failures into defers for
4588 both kinds of timeout. A transport filter timeout is now identified in
4591 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4592 systems where "make" and "gmake" are different, calling "gmake" at top
4593 level broke things. I've arranged for the value of $(MAKE) to be passed
4594 from the Makefile to this script so that it can call the same version of
4598 A note about Exim versions 4.44 and 4.50
4599 ----------------------------------------
4601 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4602 changes of various kinds. As a consequence, a big documentation update was
4603 needed. This delayed the release for rather longer than seemed good, especially
4604 in the light of a couple of (minor) security issues. Therefore, the changes
4605 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4606 release. So 4.44 and 4.50 are in effect two different branches that both start
4609 I have left the 4.50 change log unchanged; it contains all the changes since
4610 4.43. The change log for 4.44 is below; many of its items are identical to
4611 those for 4.50. This seems to be the most sensible way to preserve the
4612 historical information.
4618 1. Minor wording change to the doc/README.SIEVE file.
4620 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4621 computation of the current number of files was incorrect.
4623 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4624 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4627 4. Give more explanation in the error message when the command for a transport
4628 filter fails to execute.
4630 5. There are several places where Exim runs a non-Exim command in a
4631 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4632 was being done only for the command run by the queryprogram router. It is
4633 now done for all such subprocesses. The other cases are: ${run, transport
4634 filters, and the commands run by the lmtp and pipe transports.
4636 6. Added CONFIGURE_GROUP build-time option.
4638 7. Some older OS have a limit of 256 on the maximum number of file
4639 descriptors. Exim was using setrlimit() to set 1000 as a large value
4640 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4641 systems. I've change it so that if it can't get 1000, it tries for 256.
4643 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4644 was an oversight, and furthermore, ever since the addition of extra
4645 controls (e.g. 4.43/32), the checks on when to allow different forms of
4646 "control" were broken. There should now be diagnostics for all cases when a
4647 control that does not make sense is encountered.
4649 9. Added the /retain_sender option to "control=submission".
4651 10. $recipients is now available in the predata ACL (oversight).
4653 11. Tidy the search cache before the fork to do a delivery from a message
4654 received from the command line. Otherwise the child will trigger a lookup
4655 failure and thereby defer the delivery if it tries to use (for example) a
4656 cached ldap connection that the parent has called unbind on.
4658 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4659 of $address_data from the recipient verification was clobbered by the
4660 sender verification.
4662 13. The value of address_data from a sender verification is now available in
4663 $sender_address_data in subsequent conditions in the ACL statement.
4665 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4667 15. Added a new option "connect=<time>" to callout options, to set a different
4670 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4671 was its contents. (It was OK if the option was not defined at all.)
4673 17. A "Completed" log line is now written for messages that are removed from
4674 the spool by the -Mrm option.
4676 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4677 information about exactly what failed.
4679 19. Added -dd to debug only the daemon process.
4681 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4682 handles timeouts, both on the server side and network timeouts. Renamed the
4683 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4685 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4686 It is now set to "smtps".
4688 22. $host_address is now set to the target address during the checking of
4689 ignore_target_hosts.
4691 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4692 being passed; this would have caused $sender_host_name to have been used if
4693 matching the list had actually called for a host name (not very likely,
4694 since this list is usually IP addresses). A host name is now passed as
4697 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4698 code that specifies a non-restarting handler (typically sigaction() in
4699 modern systems) in an attempt to fix a rare and obscure crash bug.
4701 25. Narrowed the window for a race in the daemon that could cause it to ignore
4702 SIGCHLD signals. This is not a major problem, because they are used only to
4703 wake it up if nothing else does.
4705 26. A malformed maildirsize file could cause Exim to calculate negative values
4706 for the mailbox size or file count. Odd effects could occur as a result.
4707 The maildirsize information is now recalculated if the size or filecount
4710 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4711 support for a long time. Removed HAVE_SYS_VFS_H.
4713 28. Installed the latest version of exipick from John Jetmore.
4715 29. In an address list, if the pattern was not a regular expression, an empty
4716 subject address (from a bounce message) matched only if the pattern was an
4717 empty string. Non-empty patterns were not even tested. This was the wrong
4718 because it is perfectly reasonable to use an empty address as part of a
4719 database query. An empty address is now tested by patterns that are
4720 lookups. However, all the other forms of pattern expect the subject to
4721 contain a local part and a domain, and therefore, for them, an empty
4722 address still always fails if the pattern is not itself empty.
4724 30. Exim went into a mad DNS loop when attempting to do a callout where the
4725 host was specified on an smtp transport, and looking it up yielded more
4726 than one IP address.
4728 31. Re-factored the code for checking spool and log partition space into a
4729 function that finds that data and another that does the check. The former
4730 is then used to implement four new variables: $spool_space, $log_space,
4731 $spool_inodes, and $log_inodes.
4733 32. The RFC2047 encoding function was originally intended for short strings
4734 such as real names; it was not keeping to the 75-character limit for
4735 encoded words that the RFC imposes. It now respects the limit, and
4736 generates multiple encoded words if necessary. To be on the safe side, I
4737 have increased the buffer size for the ${rfc2047: expansion operator from
4740 33. It is now permitted to omit both strings after an "if" condition; if the
4741 condition is true, the result is "true". As before, when the second string
4742 is omitted, a false condition yields an empty string. This makes it less
4743 cumbersome to write custom ACL and router conditions.
4745 34. Failure to deliver a bounce message always caused it to be frozen, even if
4746 there was an errors_to setting on the router. The errors_to setting is now
4749 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4750 canonical form (fully expanded) before being placed in
4751 $sender_host_address.
4753 36. The table in the code that translates DNS record types into text (T_A to
4754 "A" for instance) was missing entries for NS and CNAME. It is just possible
4755 that this could have caused confusion if both these types were looked up
4756 for the same domain, because the text type is used as part of Exim's
4757 per-process caching. But the chance of anyone hitting this buglet seems
4760 37. The dnsdb lookup has been extended in a number of ways.
4762 (1) There is a new type, "zns", which walks up the domain tree until it
4763 finds some nameserver records. It should be used with care.
4765 (2) There is a new type, "mxh", which is like "mx" except that it returns
4766 just the host names, not the priorities.
4768 (3) It is now possible to give a list of domains (or IP addresses) to be
4769 looked up. The behaviour when one of the lookups defers can be
4770 controlled by a keyword.
4772 (4) It is now possible to specify the separator character for use when
4773 multiple records are returned.
4775 38. The dnslists ACL condition has been extended: it is now possible to supply
4776 a list of IP addresses and/or domains to be looked up in a particular DNS
4779 39. Added log_selector=+queue_time_overall.
4781 40. When running the queue in the test harness, wait just a tad after forking a
4782 delivery process, to get repeatability of debugging output.
4784 41. Include certificate and key file names in error message when GnuTLS fails
4785 to set them up, because the GnuTLS error message doesn't include the name
4786 of the failing file when there is a problem reading it.
4788 42. Allow both -bf and -bF in the same test run.
4790 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4792 44. The "Exiscan patch" is now merged into the mainline Exim source.
4794 45. Sometimes the final signoff response after QUIT could fail to get
4795 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4796 < 0 before doing a fflush(). This bug looks as though it goes back to the
4797 introduction of TLS in release 3.20, but "sometimes" must have been rare
4798 because the tests only now provoked it.
4800 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4801 (this can affect the format of dates).
4803 47. exim_tidydb, when checking for the continued existence of a message for
4804 which it has found a message-specific retry record, was not finding
4805 messages that were in split spool directories. Consequently, it was
4806 deleting retry records that should have stayed in existence.
4808 48. Steve fixed some bugs in eximstats.
4810 49. The SPA authentication driver was not abandoning authentication and moving
4811 on to the next authenticator when an expansion was forced to fail,
4812 contradicting the general specification for all authenticators. Instead it
4813 was generating a temporary error. It now behaves as specified.
4815 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4816 (the order specifies the preference for clients). The order is now AES256,
4817 AES128, 3DES, ARCFOUR128.
4819 51. Small patch to Sieve code - explicitly set From: when generating an
4822 52. Exim crashed if a remote delivery caused a very long error message to be
4823 recorded - for instance if somebody sent an entire SpamAssassin report back
4824 as a large number of 550 error lines. This bug was coincidentally fixed by
4825 increasing the size of one of Exim's internal buffers (big_buffer) that
4826 happened as part of the Exiscan merge. However, to be on the safe side, I
4827 have made the code more robust (and fixed the comments that describe what
4830 53. Now that there can be additional text after "Completed" in log lines (if
4831 the queue_time_overall log selector is set), a one-byte patch to exigrep
4832 was needed to allow it to recognize "Completed" as not the last thing in
4835 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4836 patch that reportedly fixes this has been added. I am not expert enough to
4837 create a test for it. This is what the patch creator wrote:
4839 "I found a little strange behaviour of ldap code when working with
4840 Windows 2003 AD Domain, where users was placed in more than one
4841 Organization Units. When I tried to give exim partial DN, the exit code
4842 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4843 But simultaneously result of request was absolutely normal ldap result,
4844 so I produce this patch..."
4846 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4847 REFERENCE, so I have modified the code to exclude the patch when that macro
4850 55. Some experimental protocols are using DNS PTR records for new purposes. The
4851 keys for these records are domain names, not reversed IP addresses. The
4852 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4853 leaves it alone. Component reversal etc. now happens only for IP addresses.
4856 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4858 57. Double the size of the debug message buffer (to 2048) so that more of very
4859 long debug lines gets shown.
4861 58. The exicyclog utility now does better if the number of log files to keep
4862 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4864 59. Two changes related to the smtp_active_hostname option:
4866 (1) $smtp_active_hostname is now available as a variable.
4867 (2) The default for smtp_banner uses $smtp_active_hostname instead
4868 of $primary_hostname.
4870 60. The host_aton() function is supposed to be passed a string that is known
4871 to be a valid IP address. However, in the case of IPv6 addresses, it was
4872 not checking this. This is a hostage to fortune. Exim now panics and dies
4873 if the condition is not met. A case was found where this could be provoked
4874 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4875 components; fortuitously, this particular loophole had already been fixed
4876 by change 4.50/55 above.
4878 If there are any other similar loopholes, the new check in host_aton()
4879 itself should stop them being exploited. The report I received stated that
4880 data on the command line could provoke the exploit when Exim was running as
4881 exim, but did not say which command line option was involved. All I could
4882 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4883 running as the user.
4886 61. There was a buffer overflow vulnerability in the SPA authentication code
4887 (which came originally from the Samba project). I have added a test to the
4888 spa_base64_to_bits() function which I hope fixes it.
4891 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4892 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4894 63. The daemon start-up calls getloadavg() while still root for those OS that
4895 need the first call to be done as root, but it missed one case: when
4896 deliver_queue_load_max is set with deliver_drop_privilege. This is
4897 necessary for the benefit of the queue runner, because there is no re-exec
4898 when deliver_drop_privilege is set.
4900 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4901 This has been fixed.
4903 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4904 was in use, was not putting the data itself into the right store pool;
4905 consequently, it could be overwritten for a subsequent message in the same
4906 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4909 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4911 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4912 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4913 the function were treating the return as a boolean value, which happened to
4914 work because 0=false and not-0=true, but is not correct code.
4916 68. The host_aton() function was not handling scoped IPv6 addresses (those
4917 with, for example, "%eth0" on the end) correctly.
4919 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4920 negated items (that is, ~something) in unsigned ints. Some compilers
4921 apparently mutter when there is no cast.
4923 70. If an address verification called from an ACL failed, and did not produce a
4924 user-specific message (i.e. there was only a "system" message), nothing was
4925 put in $acl_verify_message. In this situation, it now puts the system
4928 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4929 session; change 4.31/43 added the unwanted input to the log line - except
4930 that it did not do this in the start of session case. It now does.
4932 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4933 This is wrong and can cause the other end to generate a synchronization
4934 error if it is another Exim or anything else that does the synchronization
4935 check. A QUIT command is no longer sent after a timeout.
4937 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4938 during host lookups.
4940 74. The defer_ok option of callout verification was not working if it was used
4941 when verifying addresses in header lines, that is, for this case:
4943 verify = header_sender/callout=defer_ok
4945 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4946 those file descriptors could be used for SMTP connections. If anything
4947 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4948 could be sent to the SMTP client, causing chaos. The daemon now opens
4949 stdin, stdout, and stderr to /dev/null when it puts itself into the
4952 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4953 log by default. The user can override this with suitable Perl magic.
4955 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4956 the log message when discard triggers, was not working for the DATA ACL or
4957 for the non-SMTP ACL.
4959 78. Error message wording change in sieve.c.
4961 79. If smtp_accept_max_per_host was set, the number of connections could be
4962 restricted to fewer than expected, because the daemon was trying to set up
4963 a new connection before checking whether the processes handling previous
4964 connections had finished. The check for completed processes is now done
4965 earlier. On busy systems, this bug wouldn't be noticed because something
4966 else would have woken the daemon, and it would have reaped the completed
4969 80. If a message was submitted locally by a user whose login name contained one
4970 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4971 It caused a spool format error. I have fixed the spool reading code. A
4972 related problem was that the "from" clause in the Received: line became
4973 illegal because of the space(s). It is now covered by ${quote_local_part.
4975 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4978 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4980 83. Patch from Sieve maintainer to fix unterminated string problem in
4981 "vacation" handling.
4983 84. Some minor changes to the Linux configuration files to help with other
4984 OS variants using glibc.
4986 85. One more patch for Sieve to update vacation handling to latest spec.
4989 ----------------------------------------------------
4990 See the note above about the 4.44 and 4.50 releases.
4991 ----------------------------------------------------
4997 1. Change 4.43/35 introduced a bug that caused file counts to be
4998 incorrectly computed when quota_filecount was set in an appendfile
5001 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5002 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5005 3. Give more explanation in the error message when the command for a transport
5006 filter fails to execute.
5008 4. There are several places where Exim runs a non-Exim command in a
5009 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5010 was being done only for the command run by the queryprogram router. It is
5011 now done for all such subprocesses. The other cases are: ${run, transport
5012 filters, and the commands run by the lmtp and pipe transports.
5014 5. Some older OS have a limit of 256 on the maximum number of file
5015 descriptors. Exim was using setrlimit() to set 1000 as a large value
5016 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5017 systems. I've change it so that if it can't get 1000, it tries for 256.
5019 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5020 was an oversight, and furthermore, ever since the addition of extra
5021 controls (e.g. 4.43/32), the checks on when to allow different forms of
5022 "control" were broken. There should now be diagnostics for all cases when a
5023 control that does not make sense is encountered.
5025 7. $recipients is now available in the predata ACL (oversight).
5027 8. Tidy the search cache before the fork to do a delivery from a message
5028 received from the command line. Otherwise the child will trigger a lookup
5029 failure and thereby defer the delivery if it tries to use (for example) a
5030 cached ldap connection that the parent has called unbind on.
5032 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5033 of $address_data from the recipient verification was clobbered by the
5034 sender verification.
5036 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5037 was its contents. (It was OK if the option was not defined at all.)
5039 11. A "Completed" log line is now written for messages that are removed from
5040 the spool by the -Mrm option.
5042 12. $host_address is now set to the target address during the checking of
5043 ignore_target_hosts.
5045 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5046 being passed; this would have caused $sender_host_name to have been used if
5047 matching the list had actually called for a host name (not very likely,
5048 since this list is usually IP addresses). A host name is now passed as
5051 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5052 code that specifies a non-restarting handler (typically sigaction() in
5053 modern systems) in an attempt to fix a rare and obscure crash bug.
5055 15. Narrowed the window for a race in the daemon that could cause it to ignore
5056 SIGCHLD signals. This is not a major problem, because they are used only to
5057 wake it up if nothing else does.
5059 16. A malformed maildirsize file could cause Exim to calculate negative values
5060 for the mailbox size or file count. Odd effects could occur as a result.
5061 The maildirsize information is now recalculated if the size or filecount
5064 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5065 support for a long time. Removed HAVE_SYS_VFS_H.
5067 18. Updated exipick to current release from John Jetmore.
5069 19. Allow an empty sender to be matched against a lookup in an address list.
5070 Previously the only cases considered were a regular expression, or an
5073 20. Exim went into a mad DNS lookup loop when doing a callout where the
5074 host was specified on the transport, if the DNS lookup yielded more than
5077 21. The RFC2047 encoding function was originally intended for short strings
5078 such as real names; it was not keeping to the 75-character limit for
5079 encoded words that the RFC imposes. It now respects the limit, and
5080 generates multiple encoded words if necessary. To be on the safe side, I
5081 have increased the buffer size for the ${rfc2047: expansion operator from
5084 22. Failure to deliver a bounce message always caused it to be frozen, even if
5085 there was an errors_to setting on the router. The errors_to setting is now
5088 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5089 canonical form (fully expanded) before being placed in
5090 $sender_host_address.
5092 24. Updated eximstats to version 1.33
5094 25. Include certificate and key file names in error message when GnuTLS fails
5095 to set them up, because the GnuTLS error message doesn't include the name
5096 of the failing file when there is a problem reading it.
5098 26. Expand error message when OpenSSL has problems setting up cert/key files.
5101 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5102 (this can affect the format of dates).
5104 28. exim_tidydb, when checking for the continued existence of a message for
5105 which it has found a message-specific retry record, was not finding
5106 messages that were in split spool directories. Consequently, it was
5107 deleting retry records that should have stayed in existence.
5109 29. eximstats updated to version 1.35
5110 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5111 1.35 - bugfix such that pie charts by volume are generated correctly
5113 30. The SPA authentication driver was not abandoning authentication and moving
5114 on to the next authenticator when an expansion was forced to fail,
5115 contradicting the general specification for all authenticators. Instead it
5116 was generating a temporary error. It now behaves as specified.
5118 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5119 (the order specifies the preference for clients). The order is now AES256,
5120 AES128, 3DES, ARCFOUR128.
5122 31. Small patch to Sieve code - explicitly set From: when generating an
5125 32. Exim crashed if a remote delivery caused a very long error message to be
5126 recorded - for instance if somebody sent an entire SpamAssassin report back
5127 as a large number of 550 error lines. This bug was coincidentally fixed by
5128 increasing the size of one of Exim's internal buffers (big_buffer) that
5129 happened as part of the Exiscan merge. However, to be on the safe side, I
5130 have made the code more robust (and fixed the comments that describe what
5133 33. Some experimental protocols are using DNS PTR records for new purposes. The
5134 keys for these records are domain names, not reversed IP addresses. The
5135 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5136 leaves it alone. Component reversal etc. now happens only for IP addresses.
5139 34. The host_aton() function is supposed to be passed a string that is known
5140 to be a valid IP address. However, in the case of IPv6 addresses, it was
5141 not checking this. This is a hostage to fortune. Exim now panics and dies
5142 if the condition is not met. A case was found where this could be provoked
5143 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5144 components; fortuitously, this particular loophole had already been fixed
5145 by change 4.50/55 or 4.44/33 above.
5147 If there are any other similar loopholes, the new check in host_aton()
5148 itself should stop them being exploited. The report I received stated that
5149 data on the command line could provoke the exploit when Exim was running as
5150 exim, but did not say which command line option was involved. All I could
5151 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5152 running as the user.
5155 35. There was a buffer overflow vulnerability in the SPA authentication code
5156 (which came originally from the Samba project). I have added a test to the
5157 spa_base64_to_bits() function which I hope fixes it.
5160 36. The daemon start-up calls getloadavg() while still root for those OS that
5161 need the first call to be done as root, but it missed one case: when
5162 deliver_queue_load_max is set with deliver_drop_privilege. This is
5163 necessary for the benefit of the queue runner, because there is no re-exec
5164 when deliver_drop_privilege is set.
5166 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5167 was in use, was not putting the data itself into the right store pool;
5168 consequently, it could be overwritten for a subsequent message in the same
5169 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5172 38. Sometimes the final signoff response after QUIT could fail to get
5173 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5174 < 0 before doing a fflush(). This bug looks as though it goes back to the
5175 introduction of TLS in release 3.20, but "sometimes" must have been rare
5176 because the tests only now provoked it.
5182 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5183 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5184 Now it returns an integer. A number of calls were still expecting a T/F
5185 return. Fortuitously, in all cases, the tests worked in OK situations,
5186 which is the norm. However, things would have gone wrong on any write
5187 failures on the smtp file descriptor. This function is used when sending
5188 messages over SMTP and also when doing verify callouts.
5190 2. When Exim is called to do synchronous delivery of a locally submitted
5191 message (the -odf or -odi options), it no longer closes stderr before doing
5194 3. Implemented the mua_wrapper option.
5196 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5198 5. Implemented the functions header_remove(), header_testname(),
5199 header_add_at_position(), and receive_remove_recipient(), and exported them
5202 6. If an ACL "warn" statement specified the addition of headers, Exim already
5203 inserted X-ACL-Warn: at the start if there was no header name. However, it
5204 was not making this test for the second and subsequent header lines if
5205 there were newlines in the string. This meant that an invalid header could
5206 be inserted if Exim was badly configured.
5208 7. Allow an ACL "warn" statement to add header lines at the start or after all
5209 the Received: headers, as well as at the end.
5211 8. Added the rcpt_4xx retry error code.
5213 9. Added postmaster_mailfrom=xxx to callout verification option.
5215 10. Added mailfrom=xxxx to the callout verification option, for verify=
5218 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5219 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5221 12. Inserted some casts to stop certain compilers warning when using pointer
5222 differences as field lengths or precisions in printf-type calls (mostly
5223 affecting debugging statements).
5225 13. Added optional readline() support for -be (dynamically loaded).
5227 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5228 same clock tick as a message's arrival, so that its received time was the
5229 same as the "first fail" time on the retry record, and that message
5230 remained on the queue past the ultimate address timeout, every queue runner
5231 would try a delivery (because it was past the ultimate address timeout) but
5232 after another failure, the ultimate address timeout, which should have then
5233 bounced the address, did not kick in. This was a "< instead of <=" error;
5234 in most cases the first failure would have been in the next clock tick
5235 after the received time, and all would be well.
5237 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5238 being recognized when the domain list was tested by the match_domain
5239 condition in an expansion string.
5241 16. Added the ${str2b64: operator.
5243 17. Exim was always calling setrlimit() to set a large limit for the number of
5244 processes, without checking whether the existing limit was already
5245 adequate. (It did check for the limit on file descriptors.) Furthermore,
5246 errors from getrlimit() and setrlimit() were being ignored. Now they are
5247 logged to the main and panic logs, but Exim does carry on, to try to do its
5248 job under whatever limits there are.
5250 18. Imported PCRE 5.0.
5252 19. Trivial typo in log message " temporarily refused connection" (the leading
5255 20. If the log selector return_path_on_delivery was set and an address was
5256 redirected to /dev/null, the delivery process crashed because it assumed
5257 that a return path would always be set for a "successful" delivery. In this
5258 case, the whole delivery is bypassed as an optimization, and therefore no
5261 21. Internal re-arrangement: the function for sending a challenge and reading
5262 a response while authentication was assuming a zero-terminated challenge
5263 string. It's now changed to take a pointer and a length, to allow for
5264 binary data in such strings.
5266 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5268 23. Exim was not respecting finduser_retries when seeking the login of the
5269 uid under which it was called; it was always trying 10 times. (The default
5270 setting of finduser_retries is zero.) Also, it was sleeping after the final
5271 failure, which is pointless.
5273 24. Implemented tls_on_connect_ports.
5275 25. Implemented acl_smtp_predata.
5277 26. If the domain in control=submission is set empty, Exim assumes that the
5278 authenticated id is a complete email address when it generates From: or
5279 Sender: header lines.
5281 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5282 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5283 chown and chgrp in /bin and hostname in /usr/bin.
5285 28. Exim was keeping the "process log" file open after each use, just as it
5286 does for the main log. This opens the possibility of it remaining open for
5287 long periods when the USR1 signal hits a daemon. Occasional processlog
5288 errors were reported, that could have been caused by this. Anyway, it seems
5289 much more sensible not to leave this file open at all, so that is what now
5292 29. The long-running daemon process does not normally write to the log once it
5293 has entered its main loop, and it closes the log before doing so. This is
5294 so that log files can straightforwardly be renamed and moved. However,
5295 there are a couple of unusual error situations where the daemon does write
5296 log entries, and I had neglected to close the log afterwards.
5298 30. The text of an SMTP error response that was received during a remote
5299 delivery was being truncated at 512 bytes. This is too short for some of
5300 the long messages that one sometimes sees. I've increased the limit to
5303 31. It is now possible to make retry rules that apply only when a message has a
5304 specific sender, in particular, an empty sender.
5306 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5307 it possible to be selective about when SMTP synchronization is enforced.
5309 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5311 32. Implemented hosts_connection_nolog.
5313 33. Added an ACL for QUIT.
5315 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5318 35. Added mailbox_size and mailbox_filecount to appendfile.
5320 36. Added control = no_multiline_responses to ACLs.
5322 37. There was a bug in the logic of the code that waits for the clock to tick
5323 in the case where the clock went backwards by a substantial amount such
5324 that the microsecond fraction of "now" was more than the microsecond
5325 fraction of "then" (but the whole seconds number was less).
5327 38. Added support for the libradius Radius client library this is found on
5328 FreeBSD (previously only the radiusclient library was supported).
5334 1. When certain lookups returned multiple values in the form name=value, the
5335 quoting of the values was not always being done properly. Specifically:
5336 (a) If the value started with a double quote, but contained no whitespace,
5338 (b) If the value contained whitespace other than a space character (i.e.
5339 tabs or newlines or carriage returns) it was not quoted.
5340 This fix has been applied to the mysql and pgsql lookups by writing a
5341 separate quoting function and calling it from the lookup code. The fix
5342 should probably also be applied to nisplus, ibase and oracle lookups, but
5343 since I cannot test any of those, I have not disturbed their existing code.
5345 2. A hit in the callout cache for a specific address caused a log line with no
5346 reason for rejecting RCPT. Now it says "Previous (cached) callout
5347 verification failure".
5349 3. There was an off-by-one bug in the queryprogram router. An over-long
5350 return line was truncated at 256 instead of 255 characters, thereby
5351 overflowing its buffer with the terminating zero. As well as fixing this, I
5352 have increased the buffer size to 1024 (and made a note to document this).
5354 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5355 when Exim is waiting for an SMTP response from a remote server, Exim
5356 restarts its select() call on the socket, thereby resetting its timeout.
5357 This is not a problem when such interrupts are rare. Somebody set up a cron
5358 job to run exiwhat every 2 minutes, which is less than the normal select()
5359 timeout (5 or 10 minutes). This meant that the select() timeout never
5360 kicked in because it was always reset. I have fixed this by comparing the
5361 time when an interrupt arrives with the time at the start of the first call
5362 to select(). If more time than the timeout has elapsed, the interrupt is
5363 treated as a timeout.
5365 5. Some internal re-factoring in preparation for the addition of Sieve
5366 extensions (by MH). In particular, the "personal" test is moved to a
5367 separate function, and given an option for scanning Cc: and Bcc: (which is
5368 not set for Exim filters).
5370 6. When Exim created an email address using the login of the caller as the
5371 local part (e.g. when creating a From: or Sender: header line), it was not
5372 quoting the local part when it contained special characters such as @.
5374 7. Installed new OpenBSD configuration files.
5376 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5377 try to make them clearer.
5379 9. Callout options, other than the timeout value, were being ignored when
5380 verifying sender addresses in header lines. For example, when using
5382 verify = header_sender/callout=no_cache
5384 the cache was (incorrectly) being used.
5386 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5387 only the Cygwin environment.
5389 11. When return_path_on_delivery was set as a log selector, if different remote
5390 addresses in the same message used different return paths and parallel
5391 remote delivery occurred, the wrong values would sometimes be logged.
5392 (Whenever a remote delivery process finished, the return path value from
5393 the most recently started remote delivery process was logged.)
5395 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5396 lines when AUTH and/or TLS are in use. This is the "received protocol"
5397 field. Exim used to use "asmtp" for authenticated SMTP, without any
5398 indication (in the protocol name) for TLS use. Now it follows the RFC and
5399 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5400 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5401 names appear in log lines as well as in Received: header lines.
5403 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5404 extensions, and comparison tests, and to fix some bugs.
5406 14. Changes to the "personal" filter test:
5408 (1) The test was buggy in that it was just doing the equivalent of
5409 "contains" tests on header lines. For example, if a user's address was
5410 anne@some.where, the "personal" test would incorrectly be true for
5412 To: susanne@some.where
5414 This test is now done by extracting each address from the header in turn,
5415 and checking the entire address. Other tests that are part of "personal"
5416 are now done using regular expressions (for example, to check local parts
5417 of addresses in From: header lines).
5419 (2) The list of non-personal local parts in From: addresses has been
5420 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5421 taken from the Sieve specification recommendations.
5423 (3) If the message contains any header line starting with "List-" it is
5424 treated as non-personal.
5426 (4) The test for "circular" in the Subject: header line has been removed
5427 because it now seems ill-conceived.
5429 15. Minor typos in src/EDITME comments corrected.
5431 16. Installed latest exipick from John Jetmore.
5433 17. If headers_add on a router specified a text string that was too long for
5434 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5435 of string_sprintf() is now avoided.
5437 18. $message_body_size was not set (it was always zero) when running the DATA
5438 ACL and the local_scan() function.
5440 19. For the "mail" command in an Exim filter, no default was being set for
5441 the once_repeat time, causing a random time value to be used if "once" was
5442 specified. (If the value happened to be <= 0, no repeat happened.) The
5443 default is now 0s, meaning "never repeat". The "vacation" command was OK
5444 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5445 (I found it when inspecting the code).
5447 20. There is now an overall timeout for performing a callout verification. It
5448 defaults to 4 times the callout timeout, which applies to individual SMTP
5449 commands during the callout. The overall timeout applies when there is more
5450 than one host that can be tried. The timeout is checked before trying the
5451 next host. This prevents very long delays if there are a large number of
5452 hosts and all are timing out (e.g. when the network connections are timing
5453 out). The value of the overall timeout can be changed by specifying an
5454 additional sub-option for "callout", called "maxwait". For example:
5456 verify = sender/callout=5s,maxwait=20s
5458 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5459 to the end before writing, but this should make it even safer).
5461 22. Exim was forgetting that it had advertised PIPELINING for the second and
5462 subsequent messages on an SMTP connection. It was also not resetting its
5463 memory on STARTTLS and an internal HELO.
5465 23. When Exim logs an SMTP synchronization error within a session, it now
5466 records whether PIPELINING has been advertised or not.
5468 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5469 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5470 rather than long int.
5472 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5474 26. Added the never_mail option to autoreply.
5480 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5481 crash if the getsockname() call failed; this can happen if a connection is
5482 closed very soon after it is established. The problem was simply in the
5483 order in which certain operations were done, causing Exim to try to write
5484 to the SMTP stream before it had set up the file descriptor. The bug has
5485 been fixed by making things happen in the correct order.
5491 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5492 before the connection was closed, thus losing the rejection response.
5494 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5495 some early Solaris releases, but causes trouble in current releases where
5496 socklen_t is defined.
5498 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5501 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5504 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5505 that releasing the top of it at the end releases what was used for sub-
5506 expansions (unless the block got too big). However, discard this block if
5507 the first thing is a variable or header, so that we can use its block when
5508 it is dynamic (useful for very large $message_headers, for example).
5510 6. Lookups now cache *every* query, not just the most recent. A new, separate
5511 store pool is used for this. It can be recovered when all lookup caches are
5512 flushed. Lookups now release memory at the end of their result strings.
5513 This has involved some general refactoring of the lookup sources.
5515 7. Some code has been added to the store_xxx() functions to reduce the amount
5516 of flapping under certain conditions.
5518 8. log_incoming_interface used to affect only the <= reception log lines. Now
5519 it causes the local interface and port to be added to several more SMTP log
5520 lines, for example "SMTP connection from", and rejection lines.
5522 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5524 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5526 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5527 could be overwritten at the end of the current message (or the start of a
5528 new message if it was set in a HELO ACL). The value is now preserved for
5529 the duration of the SMTP connection.
5531 12. If a transport had a headers_rewrite setting, and a matching header line
5532 contained an unqualified address, that address was qualified, even if it
5533 did not match any rewriting rules. The underlying bug was that the values
5534 of the flags that permit the existence of unqualified sender and recipient
5535 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5536 non-local messages, and by -bnq for local messages) were not being
5537 preserved with the message after it was received.
5539 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5540 "next input=" as part of the text comprising the host identity instead of
5541 the correct text. The code was using the same buffer for two different
5542 strings. However, depending on which order the printing function evaluated
5543 its arguments, the bug did not always show up. Under Linux, for example, my
5544 test suite worked just fine.
5546 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5547 doesn't work with some older versions of Perl. It has been changed to "my",
5548 which in any case is probably the better facility to use.
5550 15. A really picky compiler found some instances of statements for creating
5551 error messages that either had too many or two few arguments for the format
5554 16. The size of the buffer for calls to the DNS resolver has been increased
5555 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5556 for addresses that have a lot of PTR records. This alleviates a problem; it
5557 does not fully solve it.
5559 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5560 buffer now truncates the list and logs the incident, which is the same
5561 action as happens when Exim is looking up a host name and its aliases.
5562 Previously in this situation something unpredictable would happen;
5563 sometimes it was "internal error: store_reset failed".
5565 18. If a server dropped the connection unexpectedly when an Exim client was
5566 using GnuTLS and trying to read a response, the client delivery process
5567 crashed while trying to generate an error log message.
5569 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5570 string, for example:
5572 warn message = H1: something\nH2: something
5574 the text was added as a single header line from Exim's point of view
5575 though it ended up OK in the delivered message. However, searching for the
5576 second and subsequent header lines using $h_h2: did not work. This has been
5577 fixed. Similarly, if a system filter added multiple headers in this way,
5578 the routers could not see them.
5580 20. Expanded the error message when iplsearch is called with an invalid key to
5581 suggest using net-iplsearch in a host list.
5583 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5584 ACLs are no longer actually imposed (and a message to that effect is
5587 22. If a "gecos" field in a passwd entry contained escaped characters, in
5588 particular, if it contained a \" sequence, Exim got it wrong when building
5589 a From: or a Sender: header from that name. A second bug also caused
5590 incorrect handling when an unquoted " was present following a character
5591 that needed quoting.
5593 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5594 was not being matched caselessly.
5596 24. Arranged for all hyphens in the exim.8 source to be escaped with
5599 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5600 properly. Recipient callout cache records were still being keyed to include
5601 the sender, even when use_sender was set false. This led to far more
5602 callouts that were necessary. The sender is no longer included in the key
5603 when use_sender is false.
5605 26. Added "control = submission" modifier to ACLs.
5607 27. Added the ${base62d: operator to decode base 62 numbers.
5609 28. dnsdb lookups can now access SRV records.
5611 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5612 the configuration file.
5614 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5615 (-v) mode. This makes the output for a verbose queue run more intelligible.
5617 31. Added a use_postmaster feature to recipient callouts.
5619 32. Added the $body_zerocount variable, containing the number of binary zero
5620 bytes in the message body.
5622 33. The time of last modification of the "new" subdirectory is now used as the
5623 "mailbox time last read" when there is a quota error for a maildir
5626 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5628 35. Added +ignore_unknown as a special item in host lists.
5630 36. Code for decoding IPv6 addresses in host lists is now included, even if
5631 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5632 address was recognized as an IP address, but was then not correctly decoded
5633 into binary, causing unexpected and incorrect effects when compared with
5640 1. Very minor rewording of debugging text in manualroute to say "list of
5641 hosts" instead of "hostlist".
5643 2. If verify=header_syntax was set, and a header line with an unqualified
5644 address (no domain) and a large number of spaces between the end of the
5645 name and the colon was received, the reception process suffered a buffer
5646 overflow, and (when I tested it) crashed. This was caused by some obsolete
5647 code that should have been removed. The fix is to remove it!
5649 3. When running in the test harness, delay a bit after writing a bounce
5650 message to get a bit more predictability in the log output.
5652 4. Added a call to search_tidyup() just before forking a reception process. In
5653 theory, someone could use a lookup in the expansion of smtp_accept_max_
5654 per_host which, without the tidyup, could leave open a database connection.
5656 5. Added the variables $recipient_data and $sender_data which get set from a
5657 lookup success in an ACL "recipients" or "senders" condition, or a router
5658 "senders" option, similar to $domain_data and $local_part_data.
5660 6. Moved the writing of debug_print from before to after the "senders" test
5663 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5664 problems for message scanning, either using a data ACL, or using
5665 local_scan() because the Received: header was not generated till after they
5666 were called (in order to set the time as the time of reception completion).
5667 I have revised the way this works. The header is now generated after the
5668 body is received, but before the ACL or local_scan() are called. After they
5669 are run, the timestamp in the header is updated.
5675 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5676 before starting a queue runner without re-exec. This happened only when
5677 deliver_drop_privilege was set or when the Exim user was set to root. The
5678 effect of the bug was that timeouts during subsequent deliveries caused
5679 crashes instead of being properly handled. The handler is now left at its
5680 default (and expected) setting.
5682 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5683 message, again when deliver_drop_privilege is set or Exim is run as root.
5684 The bug described in (1) was not present in this case, but the tidying up
5685 of the other signals was missing. I have made the two cases consistent.
5687 3. The ignore_target_hosts setting on a manualroute router was being ignored
5688 for hosts that were looked up using the /MX notation.
5690 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5693 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5694 operated on the sender address. After changing the $sender_address to <>
5695 for the sender address verify, Exim was re-instated it as the original
5696 (before rewriting) address, but remembering that it had rewritten it, so it
5697 wasn't rewriting it again. This bug also had the effect of breaking the
5698 sender address verification caching when the sender address was rewritten.
5700 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5701 This has been changed so that if the ip literal address matches
5702 ignore_target_hosts, the router declines.
5704 7. Added expansion conditions match_domain, match_address, and match_local_
5705 part (NOT match_host).
5707 8. The placeholder for the Received: header didn't have a length field set.
5709 9. Added code to Exim itself and to exim_lock to test for a specific race
5710 condition that could lead to file corruption when using MBX delivery. The
5711 issue is with the lockfile that is created in /tmp. If this file is removed
5712 after a process has opened it but before that process has acquired a lock,
5713 there is the potential for a second process to recreate the file and also
5714 acquire a lock. This could lead to two Exim processes writing to the file
5715 at the same time. The added code performs the same test as UW imapd; it
5716 checks after acquiring the lock that its file descriptor still refers to
5717 the same named file.
5719 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5720 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5723 11. Added the smtp_active_hostname option. If used, this will typically be made
5724 to depend on the incoming interface address. Because $interface_address is
5725 not set up until the daemon has forked a reception process, error responses
5726 that can happen earlier (such as "too many connections") no longer contain
5729 12. If an expansion in a condition on a "warn" statement fails because a lookup
5730 defers, the "warn" statement is abandoned, and the next ACL statement is
5731 processed. Previously this caused the whole ACL to be aborted.
5733 13. Added the iplsearch lookup type.
5735 14. Added ident_timeout as a log selector.
5737 15. Added tls_certificate_verified as a log selector.
5739 16. Added a global option tls_require_ciphers (compare the smtp transport
5740 option of the same name). This controls incoming TLS connections.
5742 17. I finally figured out how to make tls_require_ciphers do a similar thing
5743 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5744 before starting the TLS session.
5746 18. Tabs are now shown as \t in -bP output.
5748 19. If the log selector return_path_on_delivery was set, Exim crashed when
5749 bouncing a message because it had too many Received: header lines.
5751 20. If two routers both had headers_remove settings, and the first one included
5752 a superfluous trailing colon, the final name in the first list and the
5753 first name in the second list were incorrectly joined into one item (with a
5754 colon in the middle).
5760 1. Added -C and -D options to the exinext utility, mainly to make it easier
5761 to include in the automated testing, but these could be helpful when
5762 multiple configurations are in use.
5764 2. The exinext utility was not formatting the output nicely when there was
5765 an alternate port involved in the retry record key, nor when there was a
5766 message id as well (for retries that were specific to a specific message
5767 and a specific host). It was also confused by IPv6 addresses, because of
5768 the additional colons they contain. I have fixed the IPv4 problem, and
5769 patched it up to do a reasonable job for IPv6.
5771 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5772 delivery, the log line now contains "pipelined" if PIPELINING was used.
5774 4. An SMTP transport process used to panic and die if the bind() call to set
5775 an explicit outgoing interface failed. This has been changed; it is now
5776 treated in the same way as a connect() failure.
5778 5. A reference to $sender_host_name in the part of a conditional expansion
5779 that was being skipped was still causing a DNS lookup. This no longer
5782 6. The def: expansion condition was not recognizing references to header lines
5783 that used bh_ and bheader_.
5785 7. Added the _cache feature to named lists.
5787 8. The code for checking quota_filecount in the appendfile transport was
5788 allowing one more file than it should have been.
5790 9. For compatibility with Sendmail, the command line option
5798 and sets the incoming protocol and host name (for trusted callers). The
5799 host name and its colon can be omitted when only the protocol is to be set.
5800 Note the Exim already has two private options, -pd and -ps, that refer to
5801 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5802 "s", but I don't think that's a major issue.
5804 10. A number of refactoring changes to the code, none of which should affect
5807 (a) The number of logging options was getting close to filling up the
5808 32-bit word that was used as a bit map. I have split them into two classes:
5809 those that are passed in the argument to log_write(), and those that are
5810 only ever tested independently outside of that function. These are now in
5811 separate 32-bit words, so there is plenty of room for expansion again.
5812 There is no change in the user interface or the logging behaviour.
5814 (b) When building, for example, log lines, the code previously used a
5815 macro that called string_cat() twice, in order to add two strings. This is
5816 not really sufficiently general. Furthermore, there was one instance where
5817 it was actually wrong because one of the argument was used twice, and in
5818 one call a function was used. (As it happened, calling the function twice
5819 did not affect the overall behaviour.) The macro has been replaced by a
5820 function that can join an arbitrary number of extra strings onto a growing
5823 (c) The code for expansion conditions now uses a table and a binary chop
5824 instead of a serial search (which was left over from when there were very
5825 few conditions). Also, it now recognizes conditions like "pam" even when
5826 the relevant support is not compiled in: a suitably worded error message is
5827 given if an attempt is made to use such a condition.
5829 11. Added ${time_interval:xxxxx}.
5831 12. A bug was causing one of the ddress fields not to be passed back correctly
5832 from remote delivery subprocesses. The field in question was not being
5833 subsequently used, so this caused to problems in practice.
5835 13. Added new log selectors queue_time and deliver_time.
5837 14. Might have fixed a bug in maildirsizefile handling that threw up
5838 "unexpected character" debug warnings, and recalculated the data
5839 unnecessarily. In any case, I expanded the warning message to give more
5842 15. Added the message "Restricted characters in address" to the statements in
5843 the default ACL that block characters like @ and % in local parts.
5845 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5846 Three changes have been made:
5848 (a) There was a serious bug; a negative response to MAIL caused the whole
5849 recipient domain to be cached as invalid, thereby blocking all messages
5850 to all local parts at the same domain, from all senders. This bug has
5851 been fixed. The domain is no longer cached after a negative response to
5852 MAIL if the sender used is not empty.
5854 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5857 (c) A new callout option, "use_sender" has been added for people who want
5858 the modified behaviour.
5864 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5867 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5868 indeed breaks things for older releases.
5870 3. Added additional logging to the case where there is a problem reading data
5871 from a filter that is running in a subprocess using a pipe, in order to
5872 try to track down a specific problem.
5874 4. Testing facility fudge: when running in the test harness and attempting
5875 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5876 getting "No route to host". Convert this to a timeout.
5878 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5881 6. Some OS don't have socklen_t but use size_t instead. This affects the
5882 fifth argument of getsockopt() amongst other things. This is now
5883 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5884 can be set for individual OS. I have set it for SunOS5, OSF1, and
5885 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5886 some earlier ones do not.
5888 7. Change 4.30/15 was not doing the test caselessly.
5890 8. The standard form for an IPv6 address literal was being rejected by address
5891 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5892 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5893 this, as well as the form without the "IPv6" on the front (but only when
5894 address literals are enabled, of course).
5896 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5898 10. Exim crashed if a message with an empty sender address specified by -f
5899 encountered a router with an errors_to setting. This could be provoked only
5900 by a command such as
5904 where an empty string was supplied; "<>" did not hit this bug.
5906 11. Installed PCRE release 4.5.
5908 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5909 remained set. It is now erased.
5911 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5912 times from message ids (which are base 36 rather than the normal 62).
5914 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5915 were being counted as actual protocol errors, and logged if the log
5916 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5917 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5918 and DATA following a set of rejected RCPTs do not count as protocol errors.
5919 In other words, Exim assumes they were pipelined, though this may not
5920 actually be the case. Of course, in all cases the client gets an
5921 appropriate error code.
5923 15. If a lookup fails in an ACL condition, a message about the failure may
5924 be available; it is used if testing the ACL cannot continue, because most
5925 such messages specify what the cause of the deferral is. However, some
5926 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5927 that caused an old message to be retained and used if a later statement
5928 caused a defer, replacing the real cause of the deferral.
5930 16. If an IP address had so many PTR records that the DNS lookup buffer
5931 was not large enough to hold them, Exim could crash while trying to process
5932 the truncated data. It now detects and logs this case.
5934 17. Further to 4.21/58, another change has been made: if (and only if) the
5935 first line of a message (the first header line) ends with CRLF, a bare LF
5936 in a subsequent header line has a space inserted after it, so as not to
5937 terminate the header.
5939 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5940 unnecessarily, used atoi() instead of strtol(), and didn't check the
5941 termination when getting file sizes from file names by regex.
5943 19. Completely re-implemented the support for maildirsize files, in the light
5944 of a number of problems with the previous contributed implementation
5945 (4.30/29). In particular:
5947 . If the quota is zero, the maildirsize file is maintained, but no quota is
5950 . If the maildir directory does not exist, it is created before any attempt
5951 to write a maildirsize file.
5953 . The quota value in the file is just a cache; if the quota is changed in
5954 the transport, the new value overrides.
5956 . A regular expression is available for excluding directories from the
5959 20. The autoreply transport checks the characters in options that define the
5960 message's headers; it allows continued headers, but it was checking with
5961 isspace() after an embedded newline instead of explicitly looking for a
5964 21. If all the "regular" hosts to which an address was routed had passed their
5965 expiry times, and had not reached their retry times, the address was
5966 bounced, even if fallback hosts were defined. Now Exim should go on to try
5969 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5970 equivalent code in the SMTP transport. Some hosts send humungous responses
5971 to HELO/EHLO, more than 1024 it seems.
5973 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5974 gives compiler warnings in some environments. I've now done it "properly",
5977 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5978 (because of problems with the built-in one) was declared to return uschar *
5979 instead of char *, causing compiler failure.
5981 25. Fixed a file descriptor leak when processing alias/forward files.
5983 26. Fixed a minor format string issue in dbfn.c.
5985 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5987 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5988 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5991 29. When a sender address is verified, it is cached, to save repeating the test
5992 when there is more than one recipient in a message. However, when the
5993 verification involves a callout, it is possible for different callout
5994 options to be set for different recipients. It is too complicated to keep
5995 track of this in the cache, so now Exim always runs a verification when a
5996 callout is required, relying on the callout cache for the optimization.
5997 The overhead is duplication of the address routing, but this should not be
6000 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6001 to be verified with callout=postmaster, and the main callout worked but the
6002 postmaster check failed, the verification correctly failed. However, if a
6003 subsequent RCPT command asked for sender verification *without* the
6004 postmaster check, incorrect caching caused this verification also to fail,
6007 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6008 it was not caching the DNS options (qualify_single, search_parents) that
6009 were used when the lookup failed. A subsequent lookup with different
6010 options therefore always gave the same answer, though there were cases
6011 where it should not have. (Example: a "domains = !$mx_any" option on a
6012 dnslookup router: the "domains" option is always processed without any
6013 widening, but the router might have qualify_single set.) Now Exim uses the
6014 cached value only when the same options are set.
6016 32. Added John Jetmore's "exipick" utility to the distribution.
6018 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6019 than a timeout (e.g. a certificate is required, and is not provided), an
6020 Exim server now closes the connection immediately. Previously it waited for
6021 the client to close - but if the client is SSL, it seems that they each
6022 wait for each other, leading to a delay before one of them times out.
6024 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6025 maintained 0.8.x compatibility because I don't think many are using it, and
6026 it is clearly obsolete.
6028 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6031 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6032 client certificate was expired. A simple patch fixes this, though I don't
6033 understand the full logic of why the verify callback is called multiple
6036 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6037 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6038 which causes problems with some clients (such as the Certicom SSL Plus
6039 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6040 disables the coutermeasure allowing Eudora to connect."
6042 38. Exim was not checking that a write() to a log file succeeded. This could
6043 lead to Bad Things if a log got too big, in particular if it hit a file
6044 size limit. Exim now panics and dies if it cannot write to a log file, just
6045 as it does if it cannot open a log file.
6047 39. Modified OS/Makefile-Linux so that it now contains
6049 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6051 The two -D definitions ensure that Exim is compiled with large file
6052 support, which makes it possible to handle log files that are bigger than
6055 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6056 instance) a domain was checked against a named list that involved a lookup,
6057 causing $domain_data to be set, then another domain was checked against the
6058 same list, then the first domain was re-checked, the value of $domain_data
6059 after the final check could be wrong. In particular, if the second check
6060 failed, it could be set empty. This bug probably also applied to
6063 41. The strip_trailing_dot option was not being applied to the address given
6064 with the -f command-line option.
6066 42. The code for reading a message's header from the spool was incrementing
6067 $received_count, but never initializing it. This meant that the value was
6068 incorrect (doubled) while delivering a message in the same process in which
6069 it was received. In the most common configuration of Exim, this never
6070 happens - a fresh exec is done - but it can happen when
6071 deliver_drop_privilege is set.
6073 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6074 it now includes up to 150 characters of the unexpected data in the log
6077 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6078 and building data strings. The size of both of these buffers was 10 000
6079 bytes - far larger than anybody would *ever* want, thought I. Needless to
6080 say, somebody hit the limit. I have increased the maximum line length to
6081 20 000 and the maximum data length of concatenated lines to 100 000. I have
6082 also fixed two bugs, because there was no checking on these buffers. Tsk,
6083 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6084 buffer is too small.
6086 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6087 lsearch lookups. Now it does.
6089 46. When parsing a route_list item in a manualroute router, a fixed-length
6090 buffer was used for the list of hosts. I made this 1024 bytes long,
6091 thinking that nobody would ever have a list of hosts that long. Wrong.
6092 Somebody had a whole pile of complicated expansion conditions, and the
6093 string was silently truncated, leading to an expansion error. It turns out
6094 that it is easier to change to an unlimited length (owing to other changes
6095 that have happened since this code was originally written) than to build
6096 structure for giving a limitation error. The length of the item that
6097 expands into the list of hosts is now unlimited.
6099 47. The lsearch lookup could not handle data where the length of text line was
6100 more than 4095 characters. Such lines were truncated, leading to shortened
6101 data being returned. It should now handle lines of any length.
6103 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6104 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6107 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6108 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6109 no longer happens in comment lines. A list of replacements is now placed
6110 at the head of all of the source files, except those whose only change is
6111 to replace PERL_COMMAND in the very first #! line.
6113 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6114 messages on the queue, with a bottom-up merge sort, using code contributed
6115 by Michael Haardt. This should make operations like -bp somewhat faster on
6116 large queues. It won't affect queue runners, except when queue_run_in_order
6119 51. Installed eximstats 1.31 in the distribution.
6121 52. Added support for SRV lookups to the dnslookup router.
6123 53. If an ACL referred to $message_body or $message_body_end, the value was not
6124 reset for any messages that followed in the same SMTP session.
6126 54. The store-handling optimization for building very long strings was not
6127 differentiating between the different store pools. I don't think this
6128 actually made any difference in practice, but I've tidied it.
6130 55. While running the routers to verify a sender address, $sender_address
6131 was still set to the sender address. This is wrong, because when routing to
6132 send a bounce to the sender, it would be empty. Therefore, I have changed
6133 it so that, while verifying a sender address, $sender_address is set to <>.
6134 (There is no change to what happens when verifying a recipient address.)
6136 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6137 target A or AAAA records (if not already returned) without resetting the
6138 qualify_single or search_parents options of the DNS resolver. These are
6139 inappropriate in this case because the targets of MX and SRV records must
6140 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6141 target that, when qualified, matched something in the local domain. These
6142 two options are now turned off when doing these lookups.
6144 57. It seems that at least some releases of Reiserfs (which does not have the
6145 concept of a fixed number of inodes) returns zero and not -1 for the
6146 number of available inodes. This interacted badly with check_spool_inodes,
6147 which assumed that -1 was the "no such thing" setting. What I have done is
6148 to check that the total number of inodes is greater than zero before doing
6149 the test of how many are available.
6151 58. When a "warn" ACL statement has a log_message modifier, the message is
6152 remembered, and not repeated. This is to avoid a lot of repetition when a
6153 message has many recipients that cause the same warning to be written.
6154 However, Exim was preserving the list of already written lines for an
6155 entire SMTP session, which doesn't seem right. The memory is now reset if a
6156 new message is started.
6158 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6159 debugging output unless log_rewrite was also set.
6161 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6162 of (int)(handle) when we know that handle contains (void *)(-1).
6164 61. The Exim daemon panic-logs an error return when it closes the incoming
6165 connection. However "connection reset by peer" seems to be common, and
6166 isn't really an error worthy of noting specially, so that particular error
6169 62. When Exim is trying to find all the local interfaces, it used to panic and
6170 die if the ioctl to get the interface flags failed. However, it seems that
6171 on at least one OS (Solaris 9) it is possible to have an interface that is
6172 included in the list of interfaces, but for which you get a failure error
6173 for this call. This happens when the interface is not "plumbed" into a
6174 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6175 failure of the "get flags" call assumes that the interface is down.
6177 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6178 makes life easier for people who are doing arithmetic on fields extracted
6179 from dates, where you often get leading zeros that should not be
6180 interpreted as octal.
6182 64. Added qualify_domain to the redirect router, to override the global
6185 65. If a pathologically long header line contained very many addresses (the
6186 report of this problem mentioned 10 000) and each of them was rewritten,
6187 Exim could use up a very large amount of memory. (It kept on making new
6188 copies of the header line as it rewrote, and never released the old ones.)
6189 At the expense of a bit more processing, the header rewriting function has
6190 been changed so that it no longer eats memory in this way.
6192 66. The generation of the Received: header has been moved from the time that a
6193 message starts to be received, to the time that it finishes. The timestamp
6194 in the Received: header should now be very close to that of the <= log
6195 line. There are two side-effects of this change:
6197 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6198 logged header lines no longer include the local Received: line, because
6199 it has not yet been created. The same applies to a copy of the message
6200 that is returned to a non-SMTP sender when a message is rejected.
6202 (b) When a filter file is tested using -bf, no additional Received: header
6203 is added to the test message. After some thought, I decided that this
6206 This change does not affect the value of $received_for. It is still set
6207 after address rewriting, but before local_scan() is called.
6209 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6211 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6212 gave an unhelpful panic error message, and a defer error. I have managed to
6213 change this behaviour so that it now rejects any supplied certificate,
6214 which seems right, as the list of acceptable certificates is empty.
6216 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6217 gave an unhelpful defer error. I have not managed to make this reject any
6218 supplied certificates, but the error message it gives is "no certificate
6219 supplied", which is not helpful.
6221 70. exigrep's output now also includes lines that are not associated with any
6222 message, but which match the given pattern. Implemented by a patch from
6223 Martin Sluka, which also tidied up the Perl a bit.
6225 71. Recipient callout verification, like sender verification, was using <> in
6226 the MAIL FROM command. This isn't really the right thing, since the actual
6227 sender may affect whether the remote host accepts the recipient or not. I
6228 have changed it to use the actual sender in the callout; this means that
6229 the cache record is now keyed on a recipient/sender pair, not just the
6230 recipient address. There doesn't seem to be a real danger of callout loops,
6231 since a callout by the remote host to check the sender would use <>.
6232 [SEE ABOVE: changed after hitting problems.]
6234 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6235 temporary errors. However, in the case of such a code being given after
6236 the end of a data transmission (i.e. after ".") Exim was failing to write
6237 a retry record for the message. (Yes, there was some broken host that was
6238 actually sending 8xx at this point.)
6240 73. An unknown lookup type in a host list could cause Exim to panic-die when
6241 the list was checked. (An example that provoked this was putting <; in the
6242 middle of a list instead of at the start.) If this happened during a DATA
6243 ACL check, a -D file could be left lying around. This kind of configuration
6244 error no longer causes Exim to die; instead it causes a defer error. The
6245 incident is still logged to the main and panic logs.
6247 74. Buglet left over from Exim 3 conversion. The message "too many messages
6248 in one connection" was written to the rejectlog but not the mainlog, except
6249 when address rewriting (yes!) was being logged.
6251 75. Added write_rejectlog option.
6253 76. When a system filter was run not as root (that is, when system_filter_user
6254 was set), the values of the $n variables were not being returned to the
6255 main process; thus, they were not subsequently available in the $sn
6258 77. Added +return_path_on_delivery log selector.
6260 78. A connection timeout was being treated differently from recipients deferred
6261 when testing hosts_max_try with a message that was older than the host's
6262 retry timeout. (The host should not be counted, thus allowing all hosts to
6263 be tried at least once before bouncing.) This may have been the cause of an
6264 occasionally reported bug whereby a message would remain on the queue
6265 longer than the retry timeout, but would be bounced if a delivery was
6266 forced. I say "may" because I never totally pinned down the problem;
6267 setting up timeout/retry tests is difficult. See also the next item.
6269 79. The ultimate address timeout was not being applied to errors that involved
6270 a combination of host plus message (for example, a timeout on a MAIL
6271 command). When an address resolved to a number of possible hosts, and they
6272 were not all tried for each delivery (e.g. because of hosts_max_try), a
6273 message could remain on the queue longer than the retry timeout.
6275 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6278 81. Fixed an obscure SMTP outgoing bug which required at least the following
6279 conditions: (a) there was another message waiting for the same server;
6280 (b) the server returned 5xx to all RCPT commands in the first message so
6281 that the message was not completed; (c) the server dropped the connection
6282 or gave a negative response to the RSET that Exim sends to abort the
6283 transaction. The observed case was a dropped connection after DATA that had
6284 been sent in pipelining mode. That is, the server had advertised PIPELINING
6285 but was not implementing it correctly. The effect of the bug was incorrect
6286 behaviour, such as trying another host, and this could lead to a crash.
6292 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6293 and daemon.c were passed as pointers to ints; they should have been
6294 pointers to socklen_t variables (which are typically unsigned ints).
6296 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6299 3. Fixed a really odd bug that affected only the testing scheme; patching a
6300 certain fixed string in the binary changed the value of another string that
6301 happened to be identical to the end of the original first string.
6303 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6304 name", it returns that address as the IP address. On some operating
6305 systems (e.g. Solaris), it also passes back the IP address string as the
6306 "host name". However, on others (e.g. Linux), it passes back an empty
6307 string. Exim wasn't checking for this, and was changing the host name to an
6308 empty string, assuming it had been canonicalized.
6310 5. Although rare, it is permitted to have more than one PTR record for a given
6311 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6312 all the names associated with an address, because they do in Solaris.
6313 However, it seems that they do not in Linux for data that comes from the
6314 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6315 I found this out when I moved to a new Linux workstation and tried to run
6316 the Exim test suite.
6318 To get round this problem I have changed the code so that it now does its
6319 own call to the DNS to look up PTR records when searching for a host name.
6320 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6321 addresses that are only in /etc/hosts are still found.
6323 This behaviour is, however, controlled by an option called host_lookup_
6324 order, which defaults to "bydns:byaddr". If people want to use the other
6325 order, or indeed, just use one or the other means of lookup, they can
6326 specify it in this variable.
6328 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6329 some operating systems, this comes back from gethostbyaddr() as an empty
6330 string, and this is what Exim used to test for. However, it seems that in
6331 other systems, "." is yielded. Exim now tests for this case too.
6333 7. The values of check_spool_space and check_log_space are now held internally
6334 as a number of kilobytes instead of an absolute number of bytes. If a
6335 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6336 kilobyte. This means that much larger values can be stored.
6338 8. Exim monitor: an attempt to get the action menu when not actually pointing
6339 at a message produces an empty menu entitled "No message selected". This
6340 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6341 no entries in it ("Shell widget menu has zero width and/or height"). So I
6342 have added a single, blank menu entry in this case.
6344 9. Added ${quote_local_part.
6346 10. MIME decoding is now applied to the contents of Subject: header lines when
6349 11. Now that a reference to $sender_host_address automatically causes a reverse
6350 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6351 host lookup before query-style lookups in lists that might use this
6352 variable. This has therefore been abolished, and the "net-" prefix is no
6353 longer necessary for query-style lookups.
6355 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6356 have been a typo for LFLAGS, so it has been changed.
6358 13. The install script calls Exim with "-C /dev/null" in order to find the
6359 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6360 to be output. However, since Exim outputs its version number before the
6361 error, it didn't break the script. It just looked ugly. I fixed this by
6362 always allowing "-C /dev/null" if the caller is root.
6364 14. Ignore overlarge ACL variable number when reading spool file - insurance
6365 against a later release with more variables having written the file.
6367 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6368 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6369 this, as well as the form without the "IPv6" on the front.
6371 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6372 OS/Makefile-Darwin file.
6374 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6375 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6377 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6378 variable. However, some OS use ints for inodes. I've added cast to long int
6379 to get rid of the compiler warning.
6381 19. I had forgotten to lock out "/../" in configuration file names when
6382 ALT_CONFIG_PREFIX was set.
6384 20. Routers used for verification do not need to specify transports. However,
6385 if such a router generated a host list, and callout was configured, Exim
6386 crashed, because it could not find a port number from the (non-existent)
6387 transport. It now assumes port 25 in this circumstance.
6389 21. Added the -t option to exigrep.
6391 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6392 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6393 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6394 them is set, LOOKUP_LSEARCH is forced.
6396 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6398 24. Added sender and host information to the "rejected by local_scan()" log
6399 line; previously there was no indication of these.
6401 25. Added .include_if_exists.
6403 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6404 Linux. It turns out that not all file systems support this. Apparently some
6405 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6406 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6407 not supported on the file descriptor, is now ignored when Exim is trying to
6408 sync a directory. This applies only to Linux.
6410 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6412 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6413 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6416 29. Added support for maildirsize files from supplied patch (modified a bit).
6418 30. The use of :fail: followed by an empty string could lead Exim to respond to
6419 sender verification failures with (e.g.):
6421 550 Verification failed for <xxx>
6422 550 Sender verify failed
6424 where the first response line was missing the '-' that indicates it is not
6425 the final line of the response.
6427 31. The loop for finding the name of the user that called Exim had a hardwired
6428 limit of 10; it now uses the value of finduser_retries, which is used for
6429 all other user lookups.
6431 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6434 33. Exim was neglecting to zero errno before one call of strtol() when
6435 expanding a string and expecting an integer value. On some systems this
6436 resulted in spurious "integer overflow" errors. Also, it was casting the
6437 result into an int without checking.
6439 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6440 did not work. The code looks as if it has *never* worked, though it appears
6441 to have been documented since at least release 1.62. I have made it work.
6443 35. The "timeout_DNS" error in retry rules, also documented since at least
6444 1.62, also never worked. As it isn't clear exactly what this means, and
6445 clearly it isn't a major issue, I have abolished the feature by treating it
6446 as "timeout", and writing a warning to the main and panic logs.
6448 36. The display of retry rules for -brt wasn't always showing the error code
6451 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6452 timeout_connect_A, timeout_connect_MX.
6454 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6455 to the empty sender.
6457 39. The daemon was not analysing the content of -oX till after it had closed
6458 stderr and disconnected from the controlling terminal. This meant that any
6459 syntax errors were only noted on the panic log, and the return code from
6460 the command was 0. By re-arranging the code a little, I've made the
6461 decoding happen first, so such errors now appear on stderr, and the return
6462 code is 1. However, the actual setting up of the sockets still happens in
6463 the disconnected process, so errors there are still only recorded on the
6466 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6467 connections (as happens on some IP stacks) was logged at start up time as
6468 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6469 it from "IPv6 and IPv4", which means that two separate sockets are being
6472 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6473 says whether AF_INET or AF_INET6 was passed as an argument.
6475 42. Exiwhat output was messed up when time zones were included in log
6478 43. Exiwhat now gives more information about the daemon's listening ports,
6479 and whether -tls-on-connect was used.
6481 44. The "port" option of the smtp transport is now expanded.
6483 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6484 silently ignored. Now an error message is written to the main and panic
6487 46. There's a new ACL modifier called "logwrite" which writes to a log file
6488 as soon as it is encountered.
6490 47. Added $local_user_uid and $local_user_gid at routing time.
6492 48. Exim crashed when trying to verify a sender address that was being
6495 49. Exim was recognizing only a space character after ".include". It now also
6496 recognizes a tab character.
6498 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6499 extracting the relevant information from the specification. The man page no
6500 longer contains scrambled data for the -d option, and I've added a section
6501 at the front about calling Exim under different names.
6503 51. Added "extra_headers" argument to the "mail" command in filter files.
6505 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6508 53. Installed eximstats 1.29.
6510 54. Added transport_filter_timeout as a generic transport option.
6512 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6513 Cc: header lines. This was required by RFC 822, but it not required by RFC
6516 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6517 incoming messages that did not have them. Now it does so only if the
6518 message originates locally, that is, if there is no associated remote host
6519 address. When Resent- header lines are present, this applies to the Resent-
6520 lines rather than the non-Resent- lines.
6522 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6523 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6525 58. Messages for configuration errors now include the name of the main
6526 configuration file - useful now that there may be more than one file in a
6527 list (.included file names were always shown).
6529 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6530 for those rare installations that do not start the daemon as root or run it
6531 setuid root. I've cut out the call to initgroups() if the daemon is not
6534 60. The Exim user and group can now be bound into the binary as text strings
6535 that are looked up at the start of Exim's processing.
6537 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6539 62. Added $mailstore_basename variable.
6541 63. Installed patch to sieve.c from Michael Haardt.
6543 64. When Exim failed to open the panic log after failing to open the main log,
6544 the original message it was trying to log was written to stderr and debug
6545 output, but if they were not available (the usual case in production), it
6546 was lost. Now it is written to syslog before the two lines that record the
6547 failures to open the logs.
6549 65. Users' Exim filters run in subprocesses under the user's uid. It is
6550 possible for a "deliver" command or an alias in a "personal" command to
6551 provoke an address rewrite. If logging of address rewriting is configured,
6552 this fails because the process is not running as root or exim. There may be
6553 a better way of dealing with this, but for the moment (because 4.30 needs
6554 to be released), I have disabled address rewrite logging when running a
6555 filter in a non-root, non-exim process.
6561 1. The buildconfig auxiliary program wasn't quoting the value set for
6562 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6563 not defined. This bug was masked in 4.22 by the effect that was fixed in
6566 2. Some messages that were rejected after a message id was allocated were
6567 shown as "incomplete" by exigrep. It no longer does this for messages that
6568 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6570 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6571 have allow_domain_literals set, the ID did not get logged in the <= line.
6572 Domain literals are now always recognized in Message-ID: header lines.
6574 4. The first argument for a ${extract expansion item is the key name or field
6575 number. Leading and trailing spaces in this item were not being ignored,
6576 causing some misleading effects.
6578 5. When deliver_drop_privilege was set, single queue runner processes started
6579 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6580 same command in the process it spins off) were not dropping privilege.
6582 6. When the daemon running as "exim" started a queue runner, it always
6583 re-executed Exim in the spun-off process. This is a waste of effort when
6584 deliver_drop_privilege is set. The new process now just calls the
6585 queue-runner function directly.
6591 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6594 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6595 ignored. Though the use of -oP was forcing the writing of a pid file, it
6596 was always written to the default place.
6598 3. If the message "no IP address found for host xxxx" is generated during
6599 incoming verification, it is now followed by identification of the incoming
6600 connection (so you can more easily find what provoked it).
6602 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6604 5. Added some features to "harden" Exim a bit more against certain attacks:
6606 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6607 be put in Local/Makefile. This is like the never_users runtime option,
6608 but it cannot be overridden. The default setting is "root".
6610 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6611 prefix string with which any file named in a -C command line option
6614 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6615 is retained for -C and -D only if the caller of Exim is root. Without
6616 it, the exim user may also use -C and -D and retain privilege.
6618 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6619 command line option is disabled.
6621 6. Macro names set by the -D option must start with an upper case letter, just
6622 like macro names defined in the configuration file.
6624 7. Added "dereference=" facility to LDAP.
6626 8. Two instances of the typo "uknown" in the source files are fixed.
6628 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6629 the Configure-Makefile script screwed up while processing it.
6631 10. Incorporated PCRE 4.4.
6633 11. The SMTP synchronization check was not operating right at the start of an
6634 SMTP session. For example, it could not catch a HELO sent before the client
6635 waited for the greeting. There is now a check for outstanding input at the
6636 point when the greeting is written. Because of the duplex, asynchronous
6637 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6638 way, but not yet received, when the check is performed.
6640 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6641 on TCP/IP sockets, because this apparently causes some broken clients to
6644 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6645 unchanged) from the Cygwin maintainer.
6647 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6648 was supported instead of testing for mbx. Effectively a typo.
6650 15. The spa authenticator server code was not checking that the input it
6651 received was valid base64.
6653 16. The debug output line for the "set" modifier in ACLs was not showing the
6654 name of the variable that was being set.
6656 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6658 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6659 lookup on its own. Something else was needed to trigger the lookup. For
6660 example, a match in host_lookup or the need for a host name in a host list.
6661 Now, if $sender_host_name is referenced and the host name has not yet been
6662 looked up, a lookup is performed. If the lookup fails, the variable remains
6663 empty, and $host_lookup_failed is set to "1".
6665 19. Added "eqi" as a case-independent comparison operator.
6667 20. The saslauthd authentication condition could segfault if neither service
6668 nor realm was specified.
6670 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6671 error message that was logged was misleading, and incoming SMTP
6672 connections were dropped. The message is now more accurate, and temporary
6673 errors are given to SMTP connections.
6675 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6676 (except RSET) with a 421 error, until QUIT is received. However, it was
6677 failing to send a response to QUIT.
6679 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6680 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6681 ACL is now run after the helo_try_verify_hosts code.
6683 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6684 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6685 case-independent, so other case variants are also recognized). Apparently
6686 some systems use these upper case variants.
6688 25. If more than two messages were waiting for the same host, and a transport
6689 filter was specified for the transport, Exim sent two messages over the
6690 same TCP/IP connection, and then failed with "socket operation on non-
6691 socket" when it tried to send the third.
6693 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6695 27. The extern definition of crypt16() in expand.c was not being excluded when
6696 the OS had its own crypt16() function.
6698 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6699 as a preferred synonym for return_size_limit, both as an option and as an
6702 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6704 30. Changed the default configuration ACL to relax the local part checking rule
6705 for addresses that are not in any local domains. For these addresses,
6706 slashes and pipe symbols are allowed within local parts, but the sequence
6707 /../ is explicitly forbidden.
6709 31. SPA server authentication was not clearing the challenge buffer before
6712 32. log_message in a "warn" ACL statement was writing to the reject log as
6713 well as to the main log, which contradicts the documentation and doesn't
6714 seem right (because no rejection is happening). So I have stopped it.
6716 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6717 However, I am unable to do any testing of this.
6719 34. Fixed an infelicity in the appendfile transport. When checking directories
6720 for a mailbox, to see if any needed to be created, it was accidentally
6721 using path names with one or more superfluous leading slashes; tracing
6722 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6724 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6725 contents are added to the log line that is written for every discarded
6726 recipient. (Previously a log_message setting was ignored.)
6728 36. The ${quote: operator now quotes the string if it is empty.
6730 37. The install script runs exim in order to find its version number. If for
6731 some reason other than non-existence or emptiness, which it checks, it
6732 could not run './exim', it was installing it with an empty version number,
6733 i.e. as "exim-". This error state is now caught, and the installation is
6736 38. An argument was missing from the function that creates an error message
6737 when Exim fails to connect to the socket for saslauthd authentication.
6738 This could cause Exim to crash, or give a corrupted message.
6740 39. Added isip, isip4, and isip6 to ${if conditions.
6742 40. The ACL variables $acl_xx are now saved with the message, and can be
6743 accessed later in routers, transports, and filters.
6745 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6746 strings in the file are not string-expanded.
6748 42. If a MAIL command specified a SIZE value that was too large to fit into an
6749 int variable, the check against message_size_limit failed. Such values are
6750 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6751 day this will have to be increased, but I don't think I want to be around
6752 when emails are that large.
6759 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6760 iconv() is not standard in FreeBSD.
6762 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6763 IPv6 enabled. The observed symptom was a segmentation fault on return from
6764 the function os_common_find_running_interfaces() in src/os.c.
6766 3. In the check_special_case() function in daemon.c I had used "errno" as an
6767 argument name, which causes warnings on some systems. This was basically a
6768 typo, since it was named "eno" in the comments!
6770 4. The code that waits for the clock to tick (at a resolution of some fraction
6771 of a second) so as to ensure message-id uniqueness was always waiting for
6772 at least one whole tick, when it could have waited for less. [This is
6773 almost certainly not relevant at current processor speeds, where it is
6774 unlikely to ever wait at all. But we try to future-proof.]
6776 5. The function that sleeps for a time interval that includes fractions of a
6777 second contained a race. It did not block SIGALRM between setting the
6778 timer, and suspending (a couple of lines later). If the interval was short
6779 and the sigsuspend() was delayed until after it had expired, the suspension
6780 never ended. On busy systems this could lead to processes getting stuck for
6783 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6784 process, before it forks any delivery processes. The open lookup caching
6785 mechanism meant that the open file or database connection was passed into
6786 the delivery process. The problem was that delivery processes always tidy
6787 up cached lookup data. This could cause a problem for the next delivery
6788 process started by the queue runner, because the external queue runner
6789 process does not know about the closure. So the next delivery process
6790 still has data in the lookup cache. In the case of a file lookup, there was
6791 no problem because closing a file descriptor in a subprocess doesn't affect
6792 the parent. However, if the lookup was caching a connection to a database,
6793 the connection was closed, and the second delivery process was likely to
6794 see errors such as "PGSQL: query failed: server closed the connection
6795 unexpectedly". The problem has been fixed by closing all cached lookups
6796 in a queue runner before running a delivery process.
6798 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6799 seem to have the "const" qualifier which it has on other OS. I've
6802 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6803 *of the same type* (client or server) with the same public name that an
6804 error should be diagnosed.
6806 9. When Exim looked up a host name for an IP address, but failed to find the
6807 original IP address when looking up the host name (a safety check), it
6808 output the message "<ip address> does not match any IP for NULL", which was
6809 confusing, to say the least. The bug was that the host name should have
6810 appeared instead of "NULL".
6812 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6813 user that is built into the binary, and the -C or -D options is used, root
6814 privilege is dropped before the configuration file is read. In addition,
6815 logging is switched to stderr instead of the normal log files. If the
6816 configuration then re-defines the Exim user, the unprivileged environment
6817 is probably not what is expected, so Exim logs a panic warning message (but
6820 However, if deliver_drop_privilege is set, the unprivileged state may well
6821 be exactly what is intended, so the warning has been cut out in that case,
6822 and Exim is allowed to try to write to its normal log files.
6828 1. smtp_return_error_details was not giving details for temporary sender
6829 or receiver verification errors.
6831 2. Diagnose a configuration error if two authenticators have the same public
6834 3. Exim used not to create the message log file for a message until the first
6835 delivery attempt. This could be confusing when incoming messages were held
6836 for policy or load reasons. The message log file is now created at the time
6837 the message is received, and an initial "Received" line is written to it.
6839 4. The automatically generated man page for command line options had a minor
6840 bug that caused no ill effects; however, a more serious problem was that
6841 the procedure for building the man page automatically didn't always
6842 operate. Consequently, release 4.20 contains an out-of-date version. This
6843 shouldn't happen again.
6845 5. When building Exim with embedded Perl support, the script that builds the
6846 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6847 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6849 6. The freeze_tell option was not being used for messages that were frozen on
6850 arrival, either by an ACL or by local_scan().
6852 7. Added the smtp_incomplete_transaction log selector.
6854 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6855 was accepting AUTH without a new EHLO.
6857 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6858 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6861 10. Logging of TCP/IP connections (when configured) now happens in the main
6862 daemon process instead of the child process, so that the TCP/IP connection
6863 count is more accurate (but it can never be perfect).
6865 11. The use of "drop" in a nested ACL was not being handled correctly in the
6866 outer ACL. Now, if condition failure induced by the nested "drop" causes
6867 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6868 or "require"), the connection is dropped.
6870 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6871 that yield "discard" can now be used with an "accept" or a "discard" verb,
6872 but an error is generated for any others (because I can't see a useful way
6873 to define what should happen).
6875 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6876 are now processed in the same way as lines in the Exim configuration file.
6877 In particular, continuation lines are supported.
6879 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6881 15. Added -ti meaning -t -i.
6883 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6884 domains, and warn by logging if others are found.
6886 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6887 returned from GIFCONF when Exim is trying to find the list of interfaces on
6888 a host. The code in os.c has been modified to copy each ifreq to an aligned
6889 structure in all cases.
6891 Also, in some cases, the returned ifreq's were being copied to a 'struct
6892 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6893 means the last couple of bytes of an IPv6 address could be chopped if the
6894 ifreq contained only a normal sockaddr (14 bytes storage).
6896 18. Named domain lists were not supported in the hosts_treat_as_local option.
6897 An entry such as +xxxx was not recognized, and was treated as a literal
6900 19. Ensure that header lines added by a DATA ACL are included in the reject log
6901 if the ACL subsequently rejects the message.
6903 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6904 MD5 (which is deprecated).
6906 21. When testing a filter file using -bf, Exim was writing a message when it
6907 took the sender from a "From " line in the message, but it was not doing so
6908 when it took $return_path from a Return-Path: header line. It now does.
6910 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6911 with a valid header line field name (a series of printing characters
6912 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6914 23. Changed "disc" in the source to "disk" to conform to the documentation and
6915 the book and for uniformity.
6917 24. Ignore Sendmail's -Ooption=value command line item.
6919 25. When execve() failed while trying to run a command in a pipe transport,
6920 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6921 could be confused with a return value of 69 from the command itself. This
6922 has been changed to 127, the value the shell returns if it is asked to run
6923 a non-existent command. The wording for the related log line suggests a
6924 non-existent command as the problem.
6926 26. If received_header_text expands to an empty string, do not add a Received:
6927 header line to the message. (Well, it adds a token one on the spool, but
6928 marks it "old" so that it doesn't get used or transmitted.)
6930 27. Installed eximstats 1.28 (addition of -nt option).
6932 28. There was no check for failure on the call to getsockname() in the daemon
6933 code. This can fail if there is a shortage of resources on the system, with
6934 ENOMEM, for example. A temporary error is now given on failure.
6936 29. Contrary to the C standard, it seems that in some environments, the
6937 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6938 program. Exim now does this explicitly; it affects the formatting of
6939 timestamps using strftime().
6941 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6942 complaints. I've now initialized all the variables, to avoid this.
6944 32. Header lines added by a system filter were not being "seen" during
6945 transport-time rewrites.
6947 33. The info_callback() function passed to OpenSSL is set up with type void
6948 (*)(SSL *, int, int), as described somewhere. However, when calling the
6949 function (actually a macro) that sets it up, the type void(*)() is
6950 expected. I've put in a cast to prevent warnings from picky compilers.
6952 34. If a DNS black list lookup found a CNAME record, but there were no A
6953 records associated with the domain it pointed at, Exim crashed.
6955 35. If a DNS black list lookup returned more than one A record, Exim ignored
6956 all but the first. It now scans all returned addresses if a particular IP
6957 value is being sought. In this situation, the contents of the
6958 $dnslist_value variable are a list of all the addresses, separated by a
6961 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6962 to accept a host name and all its aliases if the forward lookup for any of
6963 them yielded the IP address of the incoming connection. Now it accepts only
6964 those names whose forward lookup yields the correct IP address. Any other
6965 names are discarded. This closes a loophole whereby a rogue DNS
6966 administrator could create reverse DNS records to break through a
6967 wildcarded host restriction in an ACL.
6969 37. If a user filter or a system filter that ran in a subprocess used any of
6970 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6971 the wrong values were passed to the pipe command ($thisaddress had the
6972 value of $0, $0 had the value of $1, etc). This bug was introduced by
6973 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6975 38. Improved the line breaking for long SMTP error messages from ACLs.
6976 Previously, if there was no break point between 40 and 75 characters, Exim
6977 left the rest of the message alone. Two changes have been made: (a) I've
6978 reduced the minimum length to 35 characters; (b) if it can't find a break
6979 point between 35 and 75 characters, it looks ahead and uses the first one
6980 that it finds. This may give the occasional overlong line, but at least the
6981 remaining text gets split now.
6983 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6984 file descriptors might be low, and that setting 1000 would always raise it.
6985 It turns out that in some environments, the limit is already over 1000 and
6986 that lowering it causes trouble. So now Exim takes care not to decrease it.
6988 40. When delivering a message, the value of $return_path is set to $sender_
6989 address at the start of routing (routers may change the value). By an
6990 oversight, this default was not being set up when an address was tested by
6991 -bt or -bv, which affected the outcome if any router or filter referred to
6994 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6995 log only when "message" or "log_message" are set. However, if one of the
6996 conditions was an address verification, or a call to a nested ACL, the
6997 messages generated by the underlying test were being passed through. This
6998 no longer happens. The underlying message is available in $acl_verify_
6999 message for both "message" and "log_message" expansions, so it can be
7000 passed through if needed.
7002 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7003 new expansion $bh_ to give the encoded byte string without charset
7004 translation. Translation happens only if iconv() is available; HAVE_ICONV
7005 indicates this at build time. HEADERS_CHARSET gives the charset to
7006 translate to; headers_charset can change it in the configuration, and
7007 "headers charset" can change it in an individual filter file.
7009 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7010 that creates RFC 2047 encoded "words" labels them as that charset instead
7011 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7012 expansion operator; (ii) when Exim creates a From: line for a local
7013 message; (iii) when a header line is rewritten to include a "phrase" part.
7015 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7016 buggy, causing it to skip the first lines of messages whose message ID
7017 ended in 'D'. This would not have bitten before Exim release 4.14, because
7018 message IDs were unlikely to end in 'D' before then. The effect was to have
7019 incorrect size information for certain domains.
7021 45. #include "config.h" was missing at the start of the crypt16.c module. This
7022 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7025 46. If there was a timeout during a "random" callout check, Exim treated it as
7026 a failure of the random address, and carried on sending RSET and the real
7027 address. If the delay was just some slowness somewhere, the response to the
7028 original RCPT would be taken as a response to RSET and so on, causing
7029 mayhem of various kinds.
7031 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7032 when I implemented it. It didn't allow for the fact that some option values
7033 may legitimately be negative (e.g. size_addition), and it didn't even do
7034 the right test for positive values.
7036 48. Domain names in DNS records are case-independent. Exim always looks them up
7037 in lower case. Some resolvers return domain names in exactly the case they
7038 appear in the zone file, that is, they may contain uppercase letters. Not
7039 all resolvers do this - some return always lower case. Exim was treating a
7040 change of case by a resolver as a change of domain, similar to a widening
7041 of a domain abbreviation. This triggered its re-routing code and so it was
7042 trying to route what was effectively the same domain again. This normally
7043 caused routing to fail (because the router wouldn't handle the domain
7044 twice). Now Exim checks for this case specially, and just changes the
7045 casing of the domain that it ultimately uses when it transmits the message
7048 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7051 50. If a filter generated a file delivery with a non-absolute name (possible if
7052 no home directory exists for the router), the forbid_file option was not
7055 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7056 the existing equality matching.
7058 52. Exim was using ints instead of ino_t variables in some places where it was
7059 dealing with inode numbers.
7061 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7062 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7063 called TMPDIR, and if it finds it is different, it changes its value.
7065 54. The smtp_printf() function is now made available to local_scan() so
7066 additional output lines can be written before returning. There is also an
7067 smtp_fflush() function to enable the detection of a dropped connection.
7068 The variables smtp_input and smtp_batched_input are exported to
7071 55. Changed the default runtime configuration: the message "Unknown user"
7072 has been removed from the ACL, and instead placed on the localuser router,
7073 using the cannot_route_message feature. This means that any verification
7074 failures that generate their own messages won't get overridden. Similarly,
7075 the "Unrouteable address" message that was in the ACL for unverifiable
7076 relay addresses has also been removed.
7078 56. Added hosts_avoid_esmtp to the smtp transport.
7080 57. The exicyclog script was not checking for the esoteric option
7081 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7082 will work only if exicyclog is run under the appropriate euid.
7084 58. Following a discussion on the list, the rules by which Exim recognises line
7085 endings on incoming messages have been changed. The -dropcr and drop_cr
7086 options are now no-ops, retained only for backwards compatibility. The
7087 following line terminators are recognized: LF CRLF CR. However, special
7088 processing applies to CR:
7090 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7091 nor a local message in the state where . is a terminator.
7093 (ii) If a bare CR is encountered in a header line, an extra space is added
7094 after the line terminator so as not to end the header. The reasoning
7095 behind this is that bare CRs in header lines are most likely either
7096 to be mistakes, or people trying to play silly games.
7098 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7099 was being incorrectly given as 18 bytes larger than it should have been.
7100 This is a VOB (very old bug).
7102 60. This may never have affected anything current, but just in case it has:
7103 When the local host is found other than at the start of a list of hosts,
7104 the local host, those with the same MX, and any that follow, are discarded.
7105 When the list in question was part of a longer list of hosts, the following
7106 hosts (not currently being processed) were also being discarded. This no
7107 longer happens. I'm not sure if this situation could ever has previously
7110 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7113 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7114 header. This is something that is recommended in a new Internet Draft, and
7115 is something that is documented as being done by Sendmail. There are two
7116 possible values. For messages generated by the autoreply transport, Exim
7119 Auto-Submitted: auto-replied
7121 whereas for all other generated messages (e.g. bounces) it adds
7123 Auto-Submitted: auto-generated
7125 63. The "personal" condition in filters now includes a test for the
7126 Auto-Submitted: header. If it contains the string "auto-" the message it
7127 not considered personal.
7129 64. Added rcpt_include_affixes as a generic transport option.
7131 65. Added queue_only_override (default true).
7133 66. Added the syslog_duplication option.
7135 67. If what should have been the first header line of a message consisted of
7136 a space followed by a colon, Exim was mis-interpreting it as a header line.
7137 It isn't of course - it is syntactically invalid and should therefore be
7138 treated as the start of the message body. The misbehaviour could have
7139 caused a number of strange effects, including loss of data in subsequent
7140 header lines, and spool format errors.
7142 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7143 client host had authenticated. This control can now be exercised by an ACL
7144 for more flexibility.
7146 69. By default, callouts do not happen when testing with -bh. There is now a
7147 variant, -bhc, which does actually run the callout code, including
7148 consulting and updating the callout cache.
7150 70. Added support for saslauthd authentication, courtesy of Alexander
7153 71. If statvfs() failed on the spool or log directories while checking their
7154 size for availability, Exim confusingly gave the error "space shortage".
7155 Furthermore, in debugging mode it crashed with a floating point exception.
7156 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7157 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7158 serious problem, Exim now writes to the main and panic logs when this
7159 happens, with details of the failure. It then refuses to accept the
7160 incoming message, giving the message "spool directory problem" or "log
7161 directory problem" with a 421 code for SMTP messages.
7163 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7164 0, 1, and 2 exist, because some OS complain for execs without them (see
7165 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7166 descriptors. However, the code omitted to check that the open succeeded,
7167 causing mysterious errors if for some reason the permissions on /dev/null
7168 got screwed. Now Exim writes a message to the main and panic logs, and
7169 bombs out if it can't open /dev/null.
7171 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7172 interact so that it is all more flexible. It is supposed to remain
7173 backwards compatible. Also added extra_local_interfaces.
7175 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7176 to bomb out with an assertion failure - to the client this appears as a
7177 connection drop. This problem occurs in the part of the code that was taken
7178 from the Samba project. Fortunately, the assertion is in a very simple
7179 function, so I have fixed this by reproducing the function inline in the
7180 one place where it is called, and arranging for authentication to fail
7181 instead of killing the process with assert().
7183 75. The SPA client code was not working when the server requested OEM rather
7184 than Unicode encoding.
7186 76. Added code to make require_files with a specific uid setting more usable in
7187 the case where statting the file as root fails - usually a non-root-mounted
7188 NFS file system. When this happens and the failure is EACCES, Exim now
7189 forks a subprocess and does the per-uid checking as the relevant uid.
7191 77. Added process_log_path.
7193 78. If log_file_path was not explicitly set, a setting of check_log_space or
7194 check_log_inodes was ignored.
7196 79. If a space check for the spool or log partitions fails, the incident is now
7197 logged. Of course, in the latter case the data may get lost...
7199 80. Added the %p formatting code to string_format() so that it can be used to
7200 print addresses in debug_print(). Adjusted all the address printing in the
7201 debugging in store.c to use %p rather than %d.
7203 81. There was a concern that a line of code in smtp_in.c could overflow a
7204 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7205 initially expressed, the concern was not well-founded, because trailing
7206 spaces are removed early. However, if the trailing spaces were followed by
7207 a NULL, they did not get removed, so the overflow was possible. Two fixes
7210 (a) I re-wrote the offending code in a cleaner fashion.
7211 (b) If an incoming SMTP command contains a NULL character, it is rejected
7214 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7215 runs initgroups(), so that if the Exim user is in any additional groups,
7216 they will be used during message reception.
7222 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob