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.
22 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
23 When redis_servers is set to a list of > 1 element, and the Redis servers
24 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
25 case of MOVED into a DEFER case instead, thus moving the query onto the
26 next server in the list. For a cluster of N elements, all N servers must
27 be defined in redis_servers.
29 GF/02 Catch and remove uninitialized value warning in exiqsumm
30 Check for existence of @ARGV before looking at $ARGV[0]
32 JH/01 Replace the store_release() internal interface with store_newblock(),
33 which internalises the check required to safely use the old one, plus
34 the allocate and data copy operations duplicated in both (!) of the
37 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
38 modifier. This matches the restriction on the commandline.
40 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
41 Previously only the last row was returned.
43 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
44 we assumed that tags in the header were well-formed, and parsed the
45 element content after inspecting only the first char of the tag.
46 Assumptions at that stage could crash the receive process on malformed
49 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
50 While running the DKIM ACL we operate on the Permanent memory pool so that
51 variables created with "set" persist to the DATA ACL. Also (at any time)
52 DNS lookups that fail create cache records using the Permanent pool. But
53 expansions release any allocations made on the current pool - so a dnsdb
54 lookup expansion done in the DKIM ACL releases the memory used for the
55 DNS negative-cache, and bad things result. Solution is to switch to the
56 Main pool for expansions.
57 While we're in that code, add checks on the DNS cache during store_reset,
58 active in the testsuite.
59 Problem spotted, and debugging aided, by Wolfgang Breyha.
61 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
62 When none of the hosts presented to a transport match an already-open
63 connection, close it and proceed with the list. Previously we would
64 queue the message. Spotted by Lena with Yahoo, probably involving
67 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
68 Previously a spurious "250 OK id=" response was appended to the proper
71 JH/08 The "support for" informational output now, which built with Content
72 Scanning support, has a line for the malware scanner interfaces compiled
73 in. Interface can be individually included or not at build time.
75 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
76 by the template makefile "src/EDITME". The "STREAM" support for an older
77 ClamAV interface method is removed.
79 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
80 rows affected is given instead).
82 JH/11 The runtime Berkeley DB library version is now additionally output by
83 "exim -d -bV". Previously only the compile-time version was shown.
85 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
86 SMTP connection. Previously, when one had more receipients than the
87 first, an abortive onward connection was made. Move to full support for
88 multiple onward connections in sequence, handling cutthrough connection
89 for all multi-message initiating connections.
91 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
92 routers. Previously, a multi-recipient message would fail to match the
93 onward-connection opened for the first recipient, and cause its closure.
95 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
96 a timeout on read on a GnuTLS initiating connection, resulting in the
97 initiating connection being dropped. This mattered most when the callout
98 was marked defer_ok. Fix to keep the two timeout-detection methods
101 JH/15 Relax results from ACL control request to enable cutthrough, in
102 unsupported situations, from error to silently (except under debug)
103 ignoring. This covers use with PRDR, frozen messages, queue-only and
106 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
108 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
109 metadata, resulting in a crash in free().
111 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
112 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
113 Broken also in d185889f4, with init system revamp.
115 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
116 Previously we abruptly closed the connection after reading a malware-
117 found indication; now we go on to read the "scan ok" response line,
120 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
121 ACL. Previously, a crash would result.
123 JH/19 Speed up macro lookups during configuration file read, by skipping non-
124 macro text after a replacement (previously it was only once per line) and
125 by skipping builtin macros when searching for an uppercase lead character.
127 JH/20 DANE support moved from Experimental to mainline. The Makefile control
128 for the build is renamed.
130 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
131 was allocated for every new TLS startup, meaning one per message. Fix
132 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
134 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
135 reported the original. Fix to report (as far as possible) the ACL
136 result replacing the original.
138 JH/23 Fix memory leak during multi-message connections using STARTTLS under
139 OpenSSL. Certificate information is loaded for every new TLS startup,
140 and the resources needed to be freed.
142 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
144 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
147 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
148 DATA response info to the (existing) per-recipient response info for
149 the "C=" log element. It can have useful tracking info from the
150 destination system. Patch from Simon Arlott.
152 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
153 length value. Previously this would segfault.
155 HS/02 Support Avast multiline protoocol, this allows passing flags to
156 newer versions of the scanner.
158 JH/28 Ensure that variables possibly set during message acceptance are marked
159 dead before release of memory in the daemon loop. This stops complaints
160 about them when the debug_store option is enabled. Discovered specifically
161 for sender_rate_period, but applies to a whole set of variables.
162 Do the same for the queue-runner and queue-list loops, for variables set
163 from spool message files. Do the same for the SMTP per-message loop, for
164 certain variables indirectly set in ACL operations.
166 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
167 as a multi-recipient message from a mailinglist manager). The coding had
168 an arbitrary cutoff number of characters while checking for more input;
169 enforced by writing a NUL into the buffer. This corrupted long / fast
170 input. The problem was exposed more widely when more pipelineing of SMTP
171 responses was introduced, and one Exim system was feeding another.
172 The symptom is log complaints of SMTP syntax error (NUL chars) on the
173 receiving system, and refused recipients seen by the sending system
174 (propating to people being dropped from mailing lists).
175 Discovered and pinpointed by David Carter.
177 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
178 replaced by the ${authresults } expansion.
180 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
182 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
183 allows proper process termination in container environments.
185 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
186 Previously the "final dot" had a newline after it; ensure it is CR,LF.
188 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
189 and "err_perm", deprecated since 4.83 when the RFC-defined words
190 "temperror" and "permerror" were introduced.
192 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
193 transport-filters or DKIM-signing. The restriction was lost in the
194 consolidation of verify-callout and delivery SMTP handling.
195 Extend the restriction to also cover ARC-signing.
197 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
198 in defer=pass mode supply a 450 to the initiator. Previously the message
201 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
202 tls_require_ciphers is used as before.
204 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
205 "pass_unscanned". Only tmpfails from the scanner are written to
206 the paniclog, as they may require admin intervention (permission
207 denied, license issues). Other scanner errors (like decompression
208 bombs) do not cause a paniclog entry.
210 JH/36 Fix reinitialisation of DKIM logging variable between messages.
211 Previously it was possible to log spurious information in receive log
214 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
215 triggered odd behaviour from Outlook Express clients.
217 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
220 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
221 since the IETF WG has not yet settled on that versus the original
222 "bare" representation.
224 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
225 Previously the millisecond value corrupted the output.
226 Fix also for syslog_pid=no and log_selector +pid, for which the pid
227 corrupted the output.
233 JH/01 Rework error string handling in TLS interface so that the caller in
234 more cases is responsible for logging. This permits library-sourced
235 string to be attached to addresses during delivery, and collapses
236 pairs of long lines into single ones.
238 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
239 during configuration. Wildcards are allowed and expanded.
241 JH/02 Rework error string handling in DKIM to pass more info back to callers.
242 This permits better logging.
244 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
245 do not close it down and have the child transport start it up again on
246 the passed-on TCP connection. Instead, proxy the child (and any
247 subsequent ones) for TLS via a unix-domain socket channel. Logging is
248 affected: the continued delivery log lines do not have any DNSSEC, TLS
249 Certificate or OCSP information. TLS cipher information is still logged.
251 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
252 identical IP addresses on different listening ports. Will also affect
255 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
256 add noisy ifdef guards to special-case this sillyness.
257 Patch from Bernd Kuhls.
259 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
260 than 255 are no longer allowed.
262 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
263 Disable the session-cache too, which might reduce our load. Since we
264 currrectly use a new context for every connection, both as server and
265 client, there is no benefit for these.
266 GnuTLS appears to not support tickets server-side by default (we don't
267 call gnutls_session_ticket_enable_server()) but client side is enabled
268 by default on recent versions (3.1.3 +) unless the PFS priority string
271 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
272 <https://reproducible-builds.org/specs/source-date-epoch/>.
274 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
275 the check for any unsuccessful recipients did not notice the limit, and
276 erroneously found still-pending ones.
278 JH/08 Pipeline CHUNKING command and data together, on kernels that support
279 MSG_MORE. Only in-clear (not on TLS connections).
281 JH/09 Avoid using a temporary file during transport using dkim. Unless a
282 transport-filter is involved we can buffer the headers in memory for
283 creating the signature, and read the spool data file once for the
284 signature and again for transmission.
286 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
287 4.77 as the kernel support then wasn't solid, having issues in 64bit
288 mode. Now, it's been long enough. Add support for FreeBSD also.
290 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
291 case where the routing stage had gathered several addresses to send to
292 a host before calling the transport for the first, we previously failed
293 to close down TLS in the old transport process before passing the TCP
294 connection to the new process. The new one sent a STARTTLS command
295 which naturally failed, giving a failed delivery and bloating the retry
296 database. Investigation and fix prototype from Wolfgang Breyha.
298 JH/12 Fix check on SMTP command input synchronisation. Previously there were
299 false-negatives in the check that the sender had not preempted a response
300 or prompt from Exim (running as a server), due to that code's lack of
301 awareness of the SMTP input buffering.
303 PP/04 Add commandline_checks_require_admin option.
304 Exim drops privileges sanely, various checks such as -be aren't a
305 security problem, as long as you trust local users with access to their
306 own account. When invoked by services which pass untrusted data to
307 Exim, this might be an issue. Set this option in main configuration
308 AND make fixes to the calling application, such as using `--` to stop
311 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
312 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
313 responses to those, into a single TLS record each way (this usually means
314 a single packet). As a side issue, smtp_enforce_sync now works on TLS
317 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
318 affects you only if you're dancing at the edge of the param size limits.
319 If you are, and this message makes sense to you, then: raise the
320 configured limit or use OpenSSL 1.1. Nothing we can do for older
323 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
324 cmdline element to get the documented default one. Previously it was
327 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
328 are unsuable for non-message-specific reasons - eg. connection timeout,
329 banner-time rejection.
331 JH/16 Fix logging of delivery remote port, when specified by router, under
334 PP/06 Repair manualroute's ability to take options in any order, even if one
335 is the name of a transport.
338 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
340 JH/17 Change the list-building routines interface to use the expanding-string
341 triplet model, for better allocation and copying behaviour.
343 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
344 Previously it was constructed the first time a possibly-matching string
345 was met in the configuration file input during startup; now it is done
348 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
349 compatible one, to avoid the (poorly documented) possibility of a config
350 file in the working directory redirecting the DB files, possibly correpting
351 some existing file. CVE-2017-10140 assigned for BDB.
353 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
354 cache-hot. Previously, although the result was properly cached, the
355 initial verify call returned a defer.
357 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
358 the main verify for receipient in uncached-mode.
360 JH/22 Retire historical build files to an "unsupported" subdir. These are
361 defined as "ones for which we have no current evidence of testing".
363 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
364 if present. Previously it was ignored.
366 JH/24 Start using specified-initialisers in C structure init coding. This is
367 a C99 feature (it's 2017, so now considered safe).
369 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
370 if was a fixed-sized field and bitmask ops via macros; it is now more
373 PP/07 GitHub PR 56: Apply MariaDB build fix.
374 Patch provided by Jaroslav Škarvada.
376 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
377 during Coverity cleanups [4.87 JH/47]
378 Diagnosis and fix provided by Michael Fischer v. Mollard.
380 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
381 the right size to place the terminating semicolon on its own folded
382 line, the header hash was calculated to an incorrect value thanks to
383 the (relaxed) space the fold became.
385 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
386 and confused the parent.
388 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
389 which could crash as a result. This could lead to undeliverable messages.
391 JH/28 Logging: "next input sent too soon" now shows where input was truncated
394 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
395 matters on fast-turnover and PID-randomising systems, which were getting
396 out-of-order delivery.
398 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
399 a possibly-overlapping copy. The symptom was that "Remote host closed
400 connection in response to HELO" was logged instead of the actual 4xx
403 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
404 Previously only that bufferd was discarded, resulting in SYMTP command
407 JH/32 DKIM: when a message has multiple signatures matching an identity given
408 in dkim_verify_signers, run the dkim acl once for each. Previously only
409 one run was done. Bug 2189.
411 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
412 "panic the current process" to "deliberately defer". The panic log is
413 still written with the problem list name; the mail and reject logs now
414 get a temp-reject line for the message that was being handled, saying
415 something like "domains check lookup or other defer". The SMTP 451
416 message is still "Temporary local problem".
418 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
419 A crafted sequence of BDAT commands could result in in-use memory beeing
420 freed. CVE-2017-16943.
422 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
423 from SMTP input. Previously it was always done; now only done for DATA
424 and not BDAT commands. CVE-2017-16944.
426 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
427 to the message (such as an overlong header line). Previously this was
428 not done and we did not exit BDAT mode. Followon from the previous item
429 though a different problem.
435 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
436 than -2003 did; needs libidn2 in addition to libidn.
438 JH/02 The path option on a pipe transport is now expanded before use.
440 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
441 Patch provided by "Björn", documentation fix added too.
443 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
444 missing a wire-to-host endian conversion.
446 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
447 close after a BDAT command line could be taken as a following command,
448 giving a synch failure. Fix by only checking for synch immediately
449 before acknowledging the chunk.
451 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
452 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
453 macro. Patches provided by Josh Soref.
455 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
456 Previously we did not; the RFC seems ambiguous and VRFY is not listed
457 by IANA as a service extension. However, John Klensin suggests that we
460 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
461 the dkim code may be unix-mode line endings rather than smtp wire-format
462 CRLF, so prepend a CR to any bare LF.
464 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
465 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
467 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
468 the first were themselves being wrongly included in the feed into dkim
469 processing; with most chunk sizes in use this resulted in an incorrect
470 body hash calculated value.
472 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
473 DKIM signature block, for verification. Although advised against by
474 standards it is specifically not ruled illegal.
476 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
478 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
479 missing a body hash (the bh= tag).
481 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
482 It seems that HAProxy sends the Proxy Protocol information in clear and
483 only then does a TLS startup, so do the same.
485 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
486 TCP connections (such as for Spamd) unless the daemon successfully set
487 Fast Open mode on its listening sockets. This fixes breakage seen on
488 too-old kernels or those not configured for Fast Open, at the cost of
489 requiring both directions being enabled for TFO, and TFO never being used
490 by non-daemon-related Exim processes.
492 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
493 endings, at least on the first header line. Try to canonify any that get
494 past that check, despite the cost.
496 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
497 now limited to an arbitrary five deep, while parsing addresses with the
498 strip_excess_angle_brackets option enabled.
500 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
501 instead leave the unprompted TLS handshake in socket buffer for the
502 TLS library to consume.
504 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
506 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
508 JH/16 Drop variables when they go out of scope. Memory management drops a whole
509 region in one operation, for speed, and this leaves assigned pointers
510 dangling. Add checks run only under the testsuite which checks all
511 variables at a store-reset and panics on a dangling pointer; add code
512 explicitly nulling out all the variables discovered. Fixes one known
513 bug: a transport crash, where a dangling pointer for $sending_ip_address
514 originally assigned in a verify callout, is re-used.
516 PP/06 Drop '.' from @INC in various Perl scripts.
518 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
520 PP/08 Reduce a number of compilation warnings under clang; building with
521 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
522 should be warning-free.
524 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
526 HS/01 Fix portability problems introduced by PP/08 for platforms where
527 realloc(NULL) is not equivalent to malloc() [SunOS et al].
529 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
530 chunk. This allows us to accept broken chunked messages. We need a more
531 general solution here.
533 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
534 already-broken messages in the queue.
536 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
538 JH/19 Fix reference counting bug in routing-generated-address tracking.
544 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
545 supports it and a size is available (ie. the sending peer gave us one).
547 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
548 years of being deprecated). The replacements are the ACLs
549 acl_smtp_mime and acl_not_smtp_mime.
551 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
552 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
553 or even an in-clear connection were permitted. Now, if the host lookup
554 was dnssec and dane was requested then the host is only used if the
555 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
556 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
557 if one fails this test.
558 This means that a poorly-configured remote DNS will make it incommunicado;
559 but it protects against a DNS-interception attack on it.
561 JH/04 Bug 1810: make continued-use of an open smtp transport connection
562 non-noisy when a race steals the message being considered.
564 JH/05 If main configuration option tls_certificate is unset, generate a
565 self-signed certificate for inbound TLS connections.
567 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
568 in rewrites and routers.
570 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
571 and logged a warning sing 4.83; now they are a configuration file error.
573 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
574 (lacking @domain). Apply the same qualification processing as RCPT.
576 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
578 JH/10 Support ${sha256:} applied to a string (as well as the previous
581 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
582 a cutthrough deliver is pending, as we always want to make a connection.
583 This also avoids re-routing the message when later placing the cutthrough
584 connection after a verify cache hit.
585 Do not update it with the verify result either.
587 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
588 when routing results in more than one destination address.
590 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
591 signing (which inhibits the cutthrough capability). Previously only
592 the presence of an option was tested; now an expansion evaluating as
593 empty is permissible (obviously it should depend only on data available
594 when the cutthrough connection is made).
596 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
597 the relevant preceding SMTP command did not note the pipelining mode.
599 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
600 Previously they were not counted.
602 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
603 as one having no matching records. Previously we deferred the message
604 that needed the lookup.
606 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
607 distinguished as "(=".
609 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
610 for missing MX records. Previously it only worked for missing A records.
612 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
614 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
615 after the data-go-ahead and data-ack. Patch from Jason Betts.
617 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
618 even for a "none" policy. Patch from Tony Meyer.
620 JH/22 Fix continued use of a connection for further deliveries. If a port was
621 specified by a router, it must also match for the delivery to be
624 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
625 When one of the recipients of a message was unsuitable for the connection
626 (has no matching addresses), we lost track of needing to mark it
627 deferred. As a result mail would be lost.
629 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
631 JH/25 Decoding ACL controls is now done using a binary search; the source code
632 takes up less space and should be simpler to maintain. Merge the ACL
633 condition decode tables also, with similar effect.
635 JH/26 Fix problem with one_time used on a redirect router which returned the
636 parent address unchanged. A retry would see the parent address marked as
637 delivered, so not attempt the (identical) child. As a result mail would
640 JH/27 Fix a possible security hole, wherein a process operating with the Exim
641 UID can gain a root shell. Credit to http://www.halfdog.net/ for
642 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
645 JH/28 Enable {spool,log} filesystem space and inode checks as default.
646 Main config options check_{log,spool}_{inodes,space} are now
647 100 inodes, 10MB unless set otherwise in the configuration.
649 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
650 Previously it only applied to the main-section connection policy
653 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
655 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
656 by me. Added RFC7919 DH primes as an alternative.
658 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
659 are not in the system include path.
661 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
662 GnuTLS, when a session startup failed (eg because the client disconnected)
663 Exim did stdio operations after fclose. This was exposed by a recent
664 change which nulled out the file handle after the fclose.
666 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
667 signed directly by the cert-signing cert, rather than an intermediate
668 OCSP-signing cert. This is the model used by LetsEncrypt.
670 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
672 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
673 an incoming connection.
675 HS/02 Bug 1802: Do not half-close the connection after sending a request
678 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
679 fallback to "prime256v1".
681 JH/34 SECURITY: Use proper copy of DATA command in error message.
682 Could leak key material. Remotely exploitable. CVE-2016-9963.
688 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
689 and 3.4.4 - once the server is enabled to respond to an OCSP request
690 it does even when not requested, resulting in a stapling non-aware
691 client dropping the TLS connection.
693 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
694 support variable-length bit vectors. No functional change.
696 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
697 The I= interface field on outgoing lines is now after the H= remote
698 host field, same as incoming lines. There is a separate
699 outgoing_interface log selector which allows you to disable the
702 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
703 If not running log_selector +smtp_connection the mainlog would be held
704 open indefinitely after a "too many connections" event, including to a
705 deleted file after a log rotate. Leave the per net connection logging
706 leaving it open for efficiency as that will be quickly detected by the
707 check on the next write.
709 HS/01 Bug 1671: Fix post transport crash.
710 Processing the wait-<transport> messages could crash the delivery
711 process if the message IDs didn't exist for some reason. When
712 using 'split_spool_directory=yes' the construction of the spool
713 file name failed already, exposing the same netto behaviour.
715 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
716 mime_regex ACL conditions.
718 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
719 to DSN fail messages (bounces): remote IP, remote greeting, remote response
720 to HELO, local diagnostic string.
722 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
723 log line to debug. Even when configured with a tls authenticator many
724 client connections are expected to not authenticate in this way, so
725 an authenticate fail is not an error.
727 HS/02 Add the Exim version string to the process info. This way exiwhat
728 gives some more detail about the running daemon.
730 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
731 matter for fast-change records such as DNSBLs.
733 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
734 retry record, even if constant. There may be multiple transports with
735 different interface settings and the retry behaviour needs to be kept
738 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
740 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
742 JH/10 Bug 840: fix log_defer_output option of pipe transport
744 JH/11 Bug 830: use same host for all RCPTS of a message, even under
745 hosts_randomize. This matters a lot when combined with mua_wrapper.
747 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
748 ${quote_pgsql:<string>} operator.
750 JH/13 Bug 1708: avoid misaligned access in cached lookup.
752 JH/14 Change header file name for freeradius-client. Relevant if compiling
753 with Radius support; from the Gentoo tree and checked under Fedora.
755 JH/15 Bug 1712: Introduce $prdr_requested flag variable
757 JH/16 Bug 1714: Permit an empty string as expansion result for transport
758 option transport_filter, meaning no filtering.
760 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
762 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
763 defaults to "*" (all hosts). The variable is now available when not built
764 with TLS, default unset, mainly to enable keeping the testsuite sane.
765 If a server certificate is not supplied (via tls_certificate) an error is
766 logged, and clients will find TLS connections fail on startup. Presumably
767 they will retry in-clear.
768 Packagers of Exim are strongly encouraged to create a server certificate
769 at installation time.
771 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
772 with the $config_file variable.
774 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
775 in transport context, after the attempt, and per-recipient. The latter type
776 is per host attempted. The event data is the error message, and the errno
777 information encodes the lookup type (A vs. MX) used for the (first) host,
778 and the trailing two digits of the smtp 4xx response.
780 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
781 to write to mainlog (or rejectlog, paniclog) in the window between file
782 creation and permissions/ownership being changed. Particularly affects
783 installations where exicyclog is run as root, rather than exim user;
784 result is that the running daemon panics and dies.
786 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
788 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
789 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
790 "pri" and "weight". Note that the previous implicit priority given by the
791 list order is no longer honoured.
793 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
796 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
797 by defining SUPPORT_SOCKS.
799 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
800 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
801 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
802 variables are renamed to proxy_{local,external}_{address,port}.
804 JH/27 Move Internationalisation support from Experimental to mainline, enabled
805 for a build by defining SUPPORT_I18N
807 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
808 of the query string, and make ${quote_redis:} do that quoting.
810 JH/29 Move Events support from Experimental to mainline, enabled by default
811 and removable for a build by defining DISABLE_EVENT.
813 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
815 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
816 cached by the daemon.
818 JH/32 Move Redis support from Experimental to mainline, enabled for a build
819 by defining LOOKUP_REDIS. The libhiredis library is required.
821 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
822 keys are given for lookup.
824 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
825 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
826 only supported when built with TLS support. The PolarSSL SHA routines
827 are still used when the TLS library is too old for convenient support.
829 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
830 openssl_options), for security. OpenSSL forces this from version 1.1.0
831 server-side so match that on older versions.
833 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
834 allocation for $value could be released as the expansion processing
835 concluded, but leaving the global pointer active for it.
837 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
838 and to use the domains and local_parts ACL conditions.
840 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
841 incorrectly not doubled on cutthrough transmission, hence seen as a
842 body-termination at the receiving system - resulting in truncated mails.
843 Commonly the sender saw a TCP-level error, and retransmitted the message
844 via the normal store-and-forward channel. This could result in duplicates
845 received - but deduplicating mailstores were liable to retain only the
846 initial truncated version.
848 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
850 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
852 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
853 we're in there, support oversigning also; bug 1309.
855 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
857 HS/04 Add support for keep_environment and add_environment options.
859 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
860 either intentional arithmetic overflow during PRNG, or testing config-
863 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
864 delivery resulted in actual delivery. Cancel cutthrough before DATA
867 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
868 rejects a recipient: pass the reject to the originator.
870 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
871 Many were false-positives and ignorable, but it's worth fixing the
874 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
875 for the new environment-manipulation done at startup. Move the routines
876 from being local to tls.c to being global via the os.c file.
878 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
879 an extract embedded as result-arg for a map, the first arg for extract
880 is unavailable so we cannot tell if this is a numbered or keyed
881 extraction. Accept either.
887 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
890 JH/02 The smtp transport option "multi_domain" is now expanded.
892 JH/03 The smtp transport now requests PRDR by default, if the server offers
895 JH/04 Certificate name checking on server certificates, when exim is a client,
896 is now done by default. The transport option tls_verify_cert_hostnames
897 can be used to disable this per-host. The build option
898 EXPERIMENTAL_CERTNAMES is withdrawn.
900 JH/05 The value of the tls_verify_certificates smtp transport and main options
901 default to the word "system" to access the system default CA bundle.
902 For GnuTLS, only version 3.0.20 or later.
904 JH/06 Verification of the server certificate for a TLS connection is now tried
905 (but not required) by default. The verification status is now logged by
906 default, for both outbound TLS and client-certificate supplying inbound
909 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
912 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
913 Status Notification (bounce) messages are now MIME format per RFC 3464.
914 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
915 under the control of the dsn_advertise_hosts option, and routers may
916 have a dsn_lasthop option.
918 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
919 default, modifiable by a malware= option. The list separator for
920 the options can now be changed in the usual way. Bug 68.
922 JH/10 The smtp_receive_timeout main option is now expanded before use.
924 JH/11 The incoming_interface log option now also enables logging of the
925 local interface on delivery outgoing connections.
927 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
928 if the interface and destination host and port all match.
930 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
933 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
934 Patch from Andrew Lewis.
936 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
937 now supports optional time-restrictions, weighting, and priority
938 modifiers per server. Patch originally by <rommer@active.by>.
940 JH/16 The spamd_address main option now supports a mixed list of local
941 and remote servers. Remote servers can be IPv6 addresses, and
942 specify a port-range.
944 JH/17 Bug 68: The spamd_address main option now supports an optional
945 timeout value per server.
947 JH/18 Bug 1581: Router and transport options headers_add/remove can
948 now have the list separator specified.
950 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
953 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
956 JH/21 Support for the A6 type of dns record is withdrawn.
958 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
959 rather than the verbs used.
961 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
962 from 255 to 1024 chars.
964 JH/24 Verification callouts now attempt to use TLS by default.
966 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
967 are generic router options now. The defaults didn't change.
969 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
970 Original patch from Alexander Shikoff, worked over by JH.
972 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
973 files if ZCAT_COMMAND is not executable.
975 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
977 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
979 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
980 Normally benign, it bites when the pair was led to by a CNAME;
981 modern usage is to not canonicalize the domain to a CNAME target
982 (and we were inconsistent anyway for A-only vs AAAA+A).
984 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
986 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
987 when evaluating $sender_host_dnssec.
989 JH/31 Check the HELO verification lookup for DNSSEC, adding new
990 $sender_helo_dnssec variable.
992 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
994 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
996 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
998 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
999 documented as working, but never had. Support all but $spam_report.
1001 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1002 added for tls authenticator.
1004 HS/03 Add perl_taintmode main config option
1010 TL/01 When running the test suite, the README says that variables such as
1011 no_msglog_check are global and can be placed anywhere in a specific
1012 test's script, however it was observed that placement needed to be near
1013 the beginning for it to behave that way. Changed the runtest perl
1014 script to read through the entire script once to detect and set these
1015 variables, reset to the beginning of the script, and then run through
1016 the script parsing/test process like normal.
1018 TL/02 The BSD's have an arc4random API. One of the functions to induce
1019 adding randomness was arc4random_stir(), but it has been removed in
1020 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1021 function when detected.
1023 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1024 cause callback expansion.
1026 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1027 syntax errors in an expansion can be treated as a string instead of
1028 logging or causing an error, due to the internal use of bool_lax
1029 instead of bool when processing it.
1031 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1032 server certificates when making smtp deliveries.
1034 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1036 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1038 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1040 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1041 Merged patch from Sebastian Wiedenroth.
1043 JH/05 Fix results-pipe from transport process. Several recipients, combined
1044 with certificate use, exposed issues where response data items split
1045 over buffer boundaries were not parsed properly. This eventually
1046 resulted in duplicates being sent. This issue only became common enough
1047 to notice due to the introduction of connection certificate information,
1048 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1050 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1051 size buffer was used, resulting in syntax errors when an expansion
1054 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1055 version 3.3.6 or later.
1057 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1058 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1059 both become "event_action", the variables become $event_name, $event_data
1060 and $event_defer_errno. There is a new variable $verify_mode, usable in
1061 routers, transports and related events. The tls:cert event is now also
1062 raised for inbound connections, if the main configuration event_action
1065 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1066 early OCSP support, but no stapling (appears to be less than 1.0.0).
1068 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1069 server certificate names available under the smtp transport option
1070 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1073 JH/10 Time-related extraction expansions from certificates now use the main
1074 option "timezone" setting for output formatting, and are consistent
1075 between OpenSSL and GnuTLS compilations. Bug 1541.
1077 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1078 encoded parameter in the incoming message. Bug 1558.
1080 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1081 include certificate info, eximon was claiming there were spoolfile
1084 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1086 JH/14 Log delivery-related information more consistently, using the sequence
1087 "H=<name> [<ip>]" wherever possible.
1089 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1090 are problematic for Debian distribution, omit them from the release
1093 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1095 JH/16 Fix string representation of time values on 64bit time_t architectures.
1098 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1099 output list separator was used.
1104 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1105 checkers that were complaining about end of non-void function with no
1108 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1109 This was a regression introduced in 4.83 by another bugfix.
1111 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1113 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1114 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1120 TF/01 Correctly close the server side of TLS when forking for delivery.
1122 When a message was received over SMTP with TLS, Exim failed to clear up
1123 the incoming connection properly after forking off the child process to
1124 deliver the message. In some situations the subsequent outgoing
1125 delivery connection happened to have the same fd number as the incoming
1126 connection previously had. Exim would try to use TLS and fail, logging
1127 a "Bad file descriptor" error.
1129 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1130 utilities have not been installed.
1132 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1133 temporary space as the ACL may create new global variables.
1135 TL/01 LDAP support uses per connection or global context settings, depending
1136 upon the detected version of the libraries at build time.
1138 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1139 to extract and use the src ip:port in logging and expansions as if it
1140 were a direct connection from the outside internet. PPv2 support was
1141 updated based on HAProxy spec change in May 2014.
1143 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1145 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1146 Properly escape header and check for NULL return.
1148 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1151 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1153 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1154 characters in header names, implemented as a verify condition.
1155 Contributed by Michael Fischer v. Mollard.
1157 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1158 results permerror and temperror. Previous values are deprecated but
1159 still accepted. In a future release, err_perm and err_temp will be
1160 completely removed, which will be a backward incompatibility if the
1161 ACL tests for either of these two old results. Patch contributed by
1162 user bes-internal on the mailing list.
1164 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1166 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1167 selectors, in both main and reject logs.
1169 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1172 JH/07 Add malware type "sock" for talking to simple daemon.
1174 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1176 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1177 routers/transports under cutthrough routing.
1179 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1180 numbers. Touch up "bool" conditional to keep the same definition.
1182 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1184 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1186 JH/12 Expand items in router/transport headers_add or headers_remove lists
1187 individually rather than the list as a whole. Bug 1452.
1189 Required for reasonable handling of multiple headers_ options when
1190 they may be empty; requires that headers_remove items with embedded
1191 colons must have them doubled (or the list-separator changed).
1193 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1194 view the policy declared in the DMARC record. Currently, $dmarc_status
1195 is a combined value of both the record presence and the result of the
1198 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1200 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1201 dnslookup router and the smtp transport (applying to the forward
1204 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1205 of ldap servers used for a specific lookup. Patch provided by Heiko
1208 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1209 New variable $lookup_dnssec_authenticated for observability.
1211 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1212 Patch submitted by Lars Timman.
1214 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1216 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1217 Requires trusted mode and valid format message id, aborts otherwise.
1218 Patch contributed by Heiko Schlichting.
1220 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1221 certextract with support for various fields. Bug 1358.
1223 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1224 is requested by default, modifiable by smtp transport option
1227 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1228 operate on certificate variables to give certificate fingerprints
1229 Also new ${sha256:cert_variable}.
1231 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1233 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1234 Christian Aistleitner.
1236 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1238 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1239 file. Patch from Wolfgang Breyha.
1241 JH/25 Expand the coverage of the delivery $host and $host_address to
1242 client authenticators run in verify callout. Bug 1476.
1244 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1245 align with daemon_smtp_ports. Bug 72.
1247 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1248 support and error reporting did not work properly.
1250 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1251 and is readable. Patch from Andrew Colin Kissa.
1253 TL/14 Enhance documentation of ${run expansion and how it parses the
1254 commandline after expansion, particularly in the case when an
1255 unquoted variable expansion results in an empty value.
1257 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1259 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1260 by renaming away from it. Addresses GH issue 15, reported by
1263 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1265 TL/15 SECURITY: prevent double expansion in math comparison functions
1266 (can expand unsanitized data). Not remotely exploitable.
1273 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1275 PP/02 Make -n do something, by making it not do something.
1276 When combined with -bP, the name of an option is not output.
1278 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1281 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1282 $sender_host_name and config options to manage this, and basic check
1285 PP/05 DSCP support for outbound connections and control modifier for inbound.
1287 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1288 (Only plugin which currently uses this is kerberos4, which nobody should
1289 be using, but we should make it available and other future plugins might
1290 conceivably use it, even though it would break NAT; stuff *should* be
1291 using channel bindings instead).
1293 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1294 name; added for Sendmail compatibility; requires admin caller.
1295 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1296 just ignore it); requires trusted caller.
1297 Also parse but ignore: -Ac -Am -X<logfile>
1300 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1302 TL/02 Add +smtp_confirmation as a default logging option.
1304 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1305 Patch by Magnus Holmgren from 2007-02-20.
1307 TL/04 Bugzilla 1281 - Spec typo.
1308 Bugzilla 1283 - Spec typo.
1309 Bugzilla 1290 - Spec grammar fixes.
1311 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1313 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1315 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1316 mailing list by by Dmitry Isaikin.
1318 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1320 JH/02 Support "G" suffix to numbers in ${if comparisons.
1322 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1324 NM/01 Bugzilla 1197 - Spec typo
1325 Bugzilla 1196 - Spec examples corrections
1327 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1329 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1330 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1333 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1334 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1336 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1337 "acl {{name}{arg}...}", and optional args on acl condition
1340 JH/05 Permit multiple router/transport headers_add/remove lines.
1342 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1344 JH/07 Avoid using a waiting database for a single-message-only transport.
1345 Performance patch from Paul Fisher. Bugzilla 1262.
1347 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1350 JH/09 Add $headers_added variable, with content from use of ACL modifier
1351 add_header (but not yet added to the message). Bugzilla 199.
1353 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1354 Pulled from Bugzilla 817 by Wolfgang Breyha.
1356 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1358 (nb: this is the same fix as in Exim 4.80.1)
1360 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1363 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1366 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1368 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1369 advertise SMTP AUTH mechanism to us, instead of a generic
1370 protocol violation error. Also, make Exim more robust to bad
1371 data from the Dovecot auth socket.
1373 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1375 When a queue runner is handling a message, Exim first routes the
1376 recipient addresses, during which it prunes them based on the retry
1377 hints database. After that it attempts to deliver the message to
1378 any remaining recipients. It then updates the hints database using
1381 So if a recipient address works intermittently, it can get repeatedly
1382 deferred at routing time. The retry hints record remains fresh so the
1383 address never reaches the final cutoff time.
1385 This is a fairly common occurrence when a user is bumping up against
1386 their storage quota. Exim had some logic in its local delivery code
1387 to deal with this. However it did not apply to per-recipient defers
1388 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1390 This change adds a proper retry rule check during routing so that the
1391 final cutoff time is checked against the message's age. We only do
1392 this check if there is an address retry record and there is not a
1393 domain retry record; this implies that previous attempts to handle
1394 the address had the retry_use_local_parts option turned on. We use
1395 this as an approximation for the destination being like a local
1396 delivery, as in LMTP.
1398 I suspect this new check makes the old local delivery cutoff check
1399 redundant, but I have not verified this so I left the code in place.
1401 TF/02 Correct gecos expansion when From: is a prefix of the username.
1403 Test 0254 submits a message to Exim with the header
1407 When I ran the test suite under the user fanf2, Exim expanded
1408 the header to contain my full name, whereas it should have added
1409 a Resent-Sender: header. It erroneously treats any prefix of the
1410 username as equal to the username.
1412 This change corrects that bug.
1414 GF/01 DCC debug and logging tidyup
1415 Error conditions log to paniclog rather than rejectlog.
1416 Debug lines prefixed by "DCC: " to remove any ambiguity.
1418 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1420 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1421 Bug spotted by Jeremy Harris; was flawed since initial commit.
1422 Would have resulted in OCSP responses post-SNI triggering an Exim
1423 NULL dereference and crash.
1425 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1427 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1428 Bug detection, analysis and fix by Samuel Thibault.
1429 Bugzilla 1331, Debian bug #698092.
1431 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1433 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1434 Server implementation by Todd Lyons, client by JH.
1435 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1436 config variable "prdr_enable" controls whether the server
1437 advertises the facility. If the client requests PRDR a new
1438 acl_data_smtp_prdr ACL is called once for each recipient, after
1439 the body content is received and before the acl_smtp_data ACL.
1440 The client is controlled by both of: a hosts_try_prdr option
1441 on the smtp transport, and the server advertisement.
1442 Default client logging of deliveries and rejections involving
1443 PRDR are flagged with the string "PRDR".
1445 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1446 fclose(). Diagnosis by Todd Lyons.
1448 PP/17 Update configure.default to handle IPv6 localhost better.
1449 Patch by Alain Williams (plus minor tweaks).
1452 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1453 This is now consistent with GnuTLS, and is now documented: the
1454 previous undocumented portable approach to treating the option as
1455 unset was to force an expansion failure. That still works, and
1456 an empty string is now equivalent.
1458 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1459 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1460 not performing validation itself.
1462 PP/20 Added force_command boolean option to pipe transport.
1463 Patch from Nick Koston, of cPanel Inc.
1465 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1468 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1470 PP/21 Fix eximon continuous updating with timestamped log-files.
1471 Broken in a format-string cleanup in 4.80, missed when I repaired the
1472 other false fix of the same issue.
1473 Report and fix from Heiko Schlichting.
1476 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1477 Report from Prashanth Katuri.
1479 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1480 It's SecureTransport, so affects any MacOS clients which use the
1481 system-integrated TLS libraries, including email clients.
1483 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1484 using a MIME ACL for non-SMTP local injection.
1485 Report and assistance in diagnosis by Warren Baker.
1487 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1489 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1491 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1492 last id that failed so it may be referenced in subsequent ACL's.
1494 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1497 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1498 ldap library initialization, allowing self-signed CA's to be
1499 used. Also properly sets require_cert option later in code by
1500 using NULL (global ldap config) instead of ldap handle (per
1501 session). Bug diagnosis and testing by alxgomz.
1503 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1504 the src/util/ subdirectory.
1506 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1507 renamed to Transport Post Delivery Action by Jeremy Harris, as
1510 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1511 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1512 redis_servers = needs to be configured which will be used by the redis
1513 lookup. Patch from Warren Baker, of The Packet Hub.
1515 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1517 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1518 hostname or reverse DNS when processing a host list. Used suggestions
1519 from multiple comments on this bug.
1521 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1523 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1524 Missed a few lines, added it to make the runtest require no keyboard
1527 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1528 contains upper case chars. Make router use caseful_local_part.
1530 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1531 support when GnuTLS has been built with p11-kit.
1537 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1539 This, or similar/improved, will also be change PP/11 of 4.82.
1545 PP/01 Handle short writes when writing local log-files.
1546 In practice, only affects FreeBSD (8 onwards).
1547 Bugzilla 1053, with thanks to Dmitry Isaikin.
1549 NM/01 Bugzilla 949 - Documentation tweak
1551 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1554 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1556 PP/02 Implemented gsasl authenticator.
1558 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1560 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1561 `pkg-config foo` for cflags/libs.
1563 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1564 with rest of GSASL and with heimdal_gssapi.
1566 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1567 `pkg-config foo` for cflags/libs for the TLS implementation.
1569 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1570 properties get this fed in as external SSF. A number of robustness
1571 and debugging improvements to the cyrus_sasl authenticator.
1573 PP/08 cyrus_sasl server now expands the server_realm option.
1575 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1576 Patch by Jeremy Harris.
1578 PP/10 Added dbmjz lookup type.
1580 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1582 PP/12 MAIL args handles TAB as well as SP, for better interop with
1583 non-compliant senders.
1584 Analysis and variant patch by Todd Lyons.
1586 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1587 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1588 Patch from Dirk Mueller <dmueller@suse.com>
1590 PP/13 tls_peerdn now print-escaped for spool files.
1591 Observed some $tls_peerdn in wild which contained \n, which resulted
1592 in spool file corruption.
1594 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1595 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1596 or write after TLS renegotiation, which otherwise led to messages
1599 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1600 as a tracking header (ie: a signed header comes before the signature).
1601 Patch from Wolfgang Breyha.
1603 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1604 comma-sep list; embedded commas doubled.
1606 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1608 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1610 Report and patch from Dmitry Banschikov.
1612 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1613 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1614 failures. We appear to now support TLS1.1+ with Exim.
1616 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1617 lets Exim select keys and certificates based upon TLS SNI from client.
1618 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1619 before an outbound SMTP session. New log_selector, +tls_sni.
1621 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1622 NULL dereference. Report and patch from Alun Jones.
1624 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1625 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1626 Not seeing resolver debug output on NetBSD, but suspect this is a
1627 resolver implementation change.
1629 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1630 Left warnings. Added "eximon gdb" invocation mode.
1632 PP/21 Defaulting "accept_8bitmime" to true, not false.
1634 PP/22 Added -bw for inetd wait mode support.
1636 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1637 locate the relevant includes and libraries. Made this the default.
1639 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1640 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1642 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1643 This may cause build issues on older platforms.
1645 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1646 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1647 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1648 Added SNI support via GnuTLS too.
1649 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1651 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1653 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1654 Applied second patch from Janne, implementing suggestion to default
1655 multiple-strings-in-record handling to match SPF spec.
1657 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1659 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1660 read-only, out of scope).
1661 Patch from Wolfgang Breyha, report from Stuart Northfield.
1663 PP/29 Fix three issues highlighted by clang analyser static analysis.
1664 Only crash-plausible issue would require the Cambridge-specific
1665 iplookup router and a misconfiguration.
1666 Report from Marcin Mirosław.
1668 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1670 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1671 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1672 As part of this, removing so much warning spew let me fix some minor
1673 real issues in debug logging.
1675 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1676 assignment on my part. Fixed.
1678 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1679 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1680 Janne Snabb (who went above and beyond: thank you).
1682 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1683 string otherwise requires a connection and a bunch more work and it's
1684 relatively easy to get wrong. Should also expose TLS library linkage
1687 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1688 64-bit ${eval} (JH/03).
1690 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1691 GNU libc to support some of the 64-bit stuff, should not lead to
1692 conflicts. Defined before os.h is pulled in, so if a given platform
1693 needs to override this, it can.
1695 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1696 protection layer was required, which is not implemented.
1697 Bugzilla 1254, patch from Wolfgang Breyha.
1699 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1700 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1701 tls_dhparam take prime identifiers. Also unbreak combination of
1702 OpenSSL+DH_params+TLSSNI.
1704 PP/39 Disable SSLv2 by default in OpenSSL support.
1710 PP/01 Solaris build fix for Oracle's LDAP libraries.
1711 Bugzilla 1109, patch from Stephen Usher.
1713 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1715 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1718 TF/02 Fix a couple more cases where we did not log the error message
1719 when unlink() failed. See also change 4.74-TF/03.
1721 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1722 lock up or crash if it happened to be inside a call to libc when it
1723 got a SIGUSR1 from exiwhat.
1725 The SIGUSR1 handler appends the current process status to the process
1726 log which is later printed by exiwhat. It used to use the general
1727 purpose logging code to do this, but several functions it calls are
1728 not safe for signals.
1730 The new output code in the SIGUSR1 handler is specific to the process
1731 log, and simple enough that it's easy to inspect for signal safety.
1732 Removing some special cases also simplifies the general logging code.
1733 Removing the spurious timestamps from the process log simplifies
1736 TF/04 Improved ratelimit ACL condition.
1738 The /noupdate option has been deprecated in favour of /readonly which
1739 has clearer semantics. The /leaky, /strict, and /readonly update modes
1740 are mutually exclusive. The update mode is no longer included in the
1741 database key; it just determines when the database is updated. (This
1742 means that when you upgrade Exim will forget old rate measurements.)
1744 Exim now checks that the per_* options are used with an update mode that
1745 makes sense for the current ACL. For example, when Exim is processing a
1746 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1747 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1748 must specify per_mail/readonly. If you omit the update mode it defaults to
1749 /leaky where that makes sense (as before) or /readonly where required.
1751 The /noupdate option is now undocumented but still supported for
1752 backwards compatibility. It is equivalent to /readonly except that in
1753 ACLs where /readonly is required you may specify /leaky/noupdate or
1754 /strict/noupdate which are treated the same as /readonly.
1756 A useful new feature is the /count= option. This is a generalization
1757 of the per_byte option, so that you can measure the throughput of other
1758 aggregate values. For example, the per_byte option is now equivalent
1759 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1761 The per_rcpt option has been generalized using the /count= mechanism
1762 (though it's more complicated than the per_byte equivalence). When it is
1763 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1764 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1765 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1766 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1767 Note that using per_rcpt with a non-readonly update mode in more than
1768 one ACL will cause the recipients to be double-counted. (The per_mail
1769 and per_byte options don't have this problem.)
1771 The handling of very low rates has changed slightly. If the computed rate
1772 is less than the event's count (usually one) then this event is the first
1773 after a long gap. In this case the rate is set to the same as this event's
1774 count, so that the first message of a spam run is counted properly.
1776 The major new feature is a mechanism for counting the rate of unique
1777 events. The new per_addr option counts the number of different
1778 recipients that someone has sent messages to in the last time period. It
1779 behaves like per_rcpt if all the recipient addresses are different, but
1780 duplicate recipient addresses do not increase the measured rate. Like
1781 the /count= option this is a general mechanism, so the per_addr option
1782 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1783 example, measure the rate that a client uses different sender addresses
1784 with the options per_mail/unique=$sender_address. There are further
1785 details in the main documentation.
1787 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1789 TF/06 Removed a few PCRE remnants.
1791 TF/07 Automatically extract Exim's version number from tags in the git
1792 repository when doing development or release builds.
1794 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1795 Bugzilla 879. Patch from Paul Fisher.
1797 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1798 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1801 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1803 PP/05 Variable $av_failed, true if the AV scanner deferred.
1804 Bugzilla 1078. Patch from John Horne.
1806 PP/06 Stop make process more reliably on build failure.
1807 Bugzilla 1087. Patch from Heiko Schlittermann.
1809 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1810 Bugzilla 1089. Patch from Heiko Schlittermann.
1812 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1813 Bugzilla 1131. Patch from Holger Weiß.
1815 PP/09 Handle IPv6 addresses with SPF.
1816 Bugzilla 860. Patch from Wolfgang Breyha.
1818 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1820 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1823 PP/11 match_* no longer expand right-hand-side by default.
1824 New compile-time build option, EXPAND_LISTMATCH_RHS.
1825 New expansion conditions, "inlist", "inlisti".
1827 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1829 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1831 PP/14 fix log_write() format string regression from TF/03.
1832 Bugzilla 1152. Patch from Dmitry Isaikin.
1838 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1840 PP/02 Harmonised TLS library version reporting; only show if debugging.
1841 Layout now matches that introduced for other libraries in 4.74 PP/03.
1843 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1845 PP/04 New "dns_use_edns0" global option.
1847 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1850 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1851 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1853 TK/01 Updated PolarSSL code to 0.14.2.
1854 Bugzilla 1097. Patch from Andreas Metzler.
1856 PP/07 Catch divide-by-zero in ${eval:...}.
1857 Fixes bugzilla 1102.
1859 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1862 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1863 format-string attack -- SECURITY: remote arbitrary code execution.
1865 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1866 time unintentionally subject to list matching rules, letting the header
1867 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1868 arbitrary string expansion). This allowed for information disclosure.
1870 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1871 INT_MIN/-1 -- value coerced to INT_MAX.
1877 NM/01 Workaround for PCRE version dependency in version reporting
1880 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1881 This fixes portability to compilers other than gcc, notably
1882 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1884 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1885 makefiles for portability to HP-UX and POSIX correctness.
1887 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1888 Also via indented variable definition in the Makefile.
1889 (Debugging by Oliver Heesakkers).
1891 PP/02 Restore caching of spamd results with expanded spamd_address.
1892 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1894 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1895 Improves build reliability. Fix from: Frank Elsner
1897 NM/02 Fix wide character breakage in the rfc2047 coding
1898 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1900 NM/03 Allow underscore in dnslist lookups
1901 Fixes bug 1026. Patch from Graeme Fowler
1903 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1904 Code patches from Adam Ciarcinski of NetBSD.
1906 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1909 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1910 is logged, to avoid truncation. Patch from John Horne.
1912 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1913 Patch from Jakob Hirsch.
1915 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1916 SQL string expansion failure details.
1917 Patch from Andrey Oktyabrski.
1919 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1920 Patch from Simon Arlott.
1922 PP/09 New lookups functionality failed to compile on old gcc which rejects
1923 extern declarations in function scope.
1924 Patch from Oliver Fleischmann
1926 PP/10 Use sig_atomic_t for flags set from signal handlers.
1927 Check getgroups() return and improve debugging.
1928 Fixed developed for diagnosis in bug 927 (which turned out to be
1931 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1932 Patch from Mark Zealey.
1934 PP/12 Bugzilla 1056: Improved spamd server selection.
1935 Patch from Mark Zealey.
1937 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1938 Based on patch from Heiko Schlittermann.
1940 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1941 Patch from Uwe Doering, sign-off by Michael Haardt.
1943 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1944 variable declaration deep within a block. Bug and patch from
1947 PP/15 lookups-Makefile IRIX compatibility coercion.
1949 PP/16 Make DISABLE_DKIM build knob functional.
1951 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1952 Patch by Simon Arlott
1954 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1955 variable argument macros. Our copy now differs from upstream.
1961 TF/01 Failure to get a lock on a hints database can have serious
1962 consequences so log it to the panic log.
1964 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1965 controlled using the smtp_confirmation log selector.
1967 TF/03 Include the error message when we fail to unlink a spool file.
1969 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1970 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1971 for maintaining out-of-tree patches for some time.
1973 PP/01 Bugzilla 139: Documentation and portability issues.
1974 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1975 Handle per-OS dynamic-module compilation flags.
1977 PP/02 Let /dev/null have normal permissions.
1978 The 4.73 fixes were a little too stringent and complained about the
1979 permissions on /dev/null. Exempt it from some checks.
1980 Reported by Andreas M. Kirchwitz.
1982 PP/03 Report version information for many libraries, including
1983 Exim version information for dynamically loaded libraries. Created
1984 version.h, now support a version extension string for distributors
1985 who patch heavily. Dynamic module ABI change.
1987 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1988 privilege escalation vulnerability whereby the Exim run-time user
1989 can cause root to append content of the attacker's choosing to
1992 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1995 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1996 If dropping privileges for untrusted macros, we disabled normal logging
1997 on the basis that it would fail; for the Exim run-time user, this is not
1998 the case, and it resulted in successful deliveries going unlogged.
1999 Fixed. Reported by Andreas Metzler.
2005 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2006 only prepend for the Resent-* case. Fixes regression introduced in
2007 Exim 4.70 by NM/22 for Bugzilla 607.
2009 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2010 increasing numbers of administrators be bitten by this.
2012 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2014 PP/03 Bugzilla 994: added openssl_options main configuration option.
2016 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2018 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2020 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2022 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2023 without a peer certificate, leading to a segfault because of an
2024 assumption that peers always have certificates. Be a little more
2025 paranoid. Problem reported by Martin Tscholak.
2027 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2028 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2029 NB: ClamAV planning to remove STREAM in "middle of 2010".
2030 CL also introduces -bmalware, various -d+acl logging additions and
2031 more caution in buffer sizes.
2033 PP/09 Implemented reverse_ip expansion operator.
2035 PP/10 Bugzilla 937: provide a "debug" ACL control.
2037 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2039 PP/12 Bugzilla 973: Implement --version.
2041 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2043 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2045 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2047 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2048 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2049 ignore trailing whitespace.
2051 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2053 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2056 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2057 Notification from Dr Andrew Aitchison.
2059 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2060 ExtendedDetectionInfo response format.
2061 Notification from John Horne.
2063 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2066 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2067 XSL and documented dependency on system catalogs, with examples of how
2070 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2073 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2074 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2075 configuration file which is writeable by the Exim user or group.
2077 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2078 of configuration files to cover files specified with the -C option if
2079 they are going to be used with root privileges, not just the default
2082 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2083 option (effectively making it always true).
2085 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2086 files to be used while preserving root privileges.
2088 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2089 that rogue child processes cannot use them.
2091 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2092 run-time user, instead of root.
2094 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2095 Exim run-time user without dropping privileges.
2097 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2098 result string, instead of calling string_vformat() twice with the same
2101 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2102 for other users. Others should always drop root privileges if they use
2103 -C on the command line, even for a whitelisted configure file.
2105 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2107 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2113 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2114 $data_path, and $header_path variables; fixed documentation bugs and
2117 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2118 exipick to access non-standard spools, including the "frozen" queue
2121 NM/01 Bugzilla 965: Support mysql stored procedures.
2122 Patch from Alain Williams
2124 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2126 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2127 Patch from Andreas Metzler
2129 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2130 Patch from Kirill Miazine
2132 NM/05 Bugzilla 671: Added umask to procmail example.
2134 JJ/03 installed exipick 20100323.0, fixing doc bug
2136 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2137 directory. Notification and patch from Dan Rosenberg.
2139 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2141 TK/02 Improve log output when DKIM signing operation fails.
2143 MH/01 Treat the transport option dkim_domain as a colon separated
2144 list, not as a single string, and sign the message with each element,
2145 omitting multiple occurences of the same signer.
2147 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2148 Bugzilla 985, 986. Patch by Simon Arlott
2150 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2151 Patch by Simon Arlott
2153 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2154 MBX locking. Notification from Dan Rosenberg.
2160 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2162 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2164 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2166 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2168 NM/04 Bugzilla 915: Flag broken perl installation during build.
2174 TK/01 Added patch by Johannes Berg that expands the main option
2175 "spamd_address" if it starts with a dollar sign.
2177 TK/02 Write list of recipients to X-Envelope-Sender header when building
2178 the mbox-format spool file for content scanning (suggested by Jakob
2181 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2182 (http://www.dcc-servers.net/) support via dccifd. Activated by
2183 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2185 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2186 by Mark Daniel Reidel <mr@df.eu>.
2188 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2189 When building exim an external PCRE library is now needed -
2190 PCRE is a system library on the majority of modern systems.
2191 See entry on PCRE_LIBS in EDITME file.
2193 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2194 conversation. Added nologin parameter to request.
2195 Patch contributed by Kirill Miazine.
2197 TF/01 Do not log submission mode rewrites if they do not change the address.
2199 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2201 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2202 log files in place. Contributed by Roberto Lima.
2204 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2206 TF/03 Bugzilla 615: When checking the local_parts router precondition
2207 after a local_part_suffix or local_part_prefix option, Exim now
2208 does not use the address's named list lookup cache, since this
2209 contains cached lookups for the whole local part.
2211 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2212 Robert Millan. Documentation is in experimental-spec.txt.
2214 TF/04 Bugzilla 668: Fix parallel build (make -j).
2216 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2218 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2219 Patch provided by Jan Srzednicki.
2221 TF/05 Leading white space used to be stripped from $spam_report which
2222 wrecked the formatting. Now it is preserved.
2224 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2225 that they are available at delivery time.
2227 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2229 TF/08 TLS error reporting now respects the incoming_interface and
2230 incoming_port log selectors.
2232 TF/09 Produce a more useful error message if an SMTP transport's hosts
2233 setting expands to an empty string.
2235 NM/06 Bugzilla 744: EXPN did not work under TLS.
2236 Patch provided by Phil Pennock.
2238 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2239 Patch provided by Richard Godbee.
2241 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2242 acl_smtp_notquit, added index entry.
2244 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2245 Patch provided by Eugene Bujak.
2247 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2248 accept(). Patch provided by Maxim Dounin.
2250 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2251 Patch provided by Phil Pennock.
2253 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2255 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2256 Patch provided by Brad "anomie" Jorsch.
2258 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2259 Patch provided by Dean Brooks.
2261 TK/05 Add native DKIM support (does not depend on external libraries).
2263 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2264 Patch provided by Graeme Fowler.
2266 NM/16 Bugzilla 851: Documentation example syntax fix.
2268 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2270 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2273 NM/19 Bugzilla 745: TLS version reporting.
2274 Patch provided by Phil Pennock.
2276 NM/20 Bugzilla 167: bool: condition support.
2277 Patch provided by Phil Pennock.
2279 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2280 clients. Patch provided by Phil Pennock.
2282 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2283 Patch provided by Brad "anomie" Jorsch.
2285 NM/23 Bugzilla 687: Fix misparses in eximstats.
2286 Patch provided by Heiko Schlittermann.
2288 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2289 Patch provided by Heiko Schlittermann.
2291 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2292 plus update to original patch.
2294 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2296 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2297 Patch provided by David Brownlee.
2299 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2301 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2303 NM/30 Bugzilla 888: TLS documentation bugfixes.
2305 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2307 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2308 Unlike the original bugzilla I have changed all shell scripts in src tree.
2310 NM/33 Bugzilla 898: Transport filter timeout fix.
2311 Patch by Todd Rinaldo.
2313 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2314 Patch by Serge Demonchaux.
2316 NM/35 Bugzilla 39: Base64 decode bug fixes.
2317 Patch by Jakob Hirsch.
2319 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2321 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2323 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2325 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2331 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2332 ALT-N's libdkim that I have put here:
2333 http://duncanthrax.net/exim-experimental/
2335 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2336 were called 'true' and it seems that C99 defines that as a reserved
2337 keyword to be used with 'bool' variable types. That means you could
2338 not include C99-style headers which use bools without triggering
2339 build errors in sieve.c.
2341 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2342 as mailq or other aliases. Changed the --help handling significantly
2343 to do whats expected. exim_usage() emits usage/help information.
2345 SC/01 Added the -bylocaldomain option to eximstats.
2347 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2349 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2351 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2357 PH/01 Another patch from the Sieve maintainer.
2359 PH/02 When an IPv6 address is converted to a string for single-key lookup
2360 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2361 dots are used instead of colons so that keys in lsearch files need not
2362 contain colons. This was done some time before quoting was made available
2363 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2364 (notated using the quote facility) so as to distinguish them from IPv4
2365 keys. This meant that lookups for IP addresses in host lists did not work
2366 for iplsearch lookups.
2368 This has been fixed by arranging for IPv6 addresses to be expressed with
2369 colons if the lookup type is iplsearch. This is not incompatible, because
2370 previously such lookups could never work.
2372 The situation is now rather anomalous, since one *can* have colons in
2373 ordinary lsearch keys. However, making the change in all cases is
2374 incompatible and would probably break a number of configurations.
2376 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2379 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2380 conversion specification without a maximum field width, thereby enabling
2381 a rogue spamd server to cause a buffer overflow. While nobody in their
2382 right mind would setup Exim to query an untrusted spamd server, an
2383 attacker that gains access to a server running spamd could potentially
2384 exploit this vulnerability to run arbitrary code as the Exim user.
2386 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2387 $primary_hostname instead of what libspf2 thinks the hosts name is.
2389 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2390 a directory entry by the name of the lookup key. Previously, if a
2391 symlink pointed to a non-existing file or a file in a directory that
2392 Exim lacked permissions to read, a lookup for a key matching that
2393 symlink would fail. Now it is enough that a matching directory entry
2394 exists, symlink or not. (Bugzilla 503.)
2396 PH/03 The body_linecount and body_zerocount variables are now exported in the
2399 PH/04 Added the $dnslist_matched variable.
2401 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2402 This means they are set thereafter only if the connection becomes
2405 PH/06 Added the client_condition to authenticators so that some can be skipped
2406 by clients under certain conditions.
2408 PH/07 The error message for a badly-placed control=no_multiline_responses left
2409 "_responses" off the end of the name.
2411 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2413 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2414 (without spaces) instead of just copying the configuration text.
2416 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2418 PH/11 Added $max_received_linelength.
2420 PH/12 Added +ignore_defer and +include_defer to host lists.
2422 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2423 way in which PCRE > 7.0 is built.
2425 PH/14 Implemented queue_only_load_latch.
2427 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2428 MAIL command. The effect was to mangle the value on 64-bit systems.
2430 PH/16 Another patch from the Sieve maintainer.
2432 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2434 PH/18 If a system quota error occurred while trying to create the file for
2435 a maildir delivery, the message "Mailbox is full" was not appended to the
2436 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2437 only to a quota excession during the actual writing of the file.
2439 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2440 characters?) which causes problems in log lines. The DN values are now
2441 passed through string_printing() before being added to log lines.
2443 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2444 and InterBase are left for another time.)
2446 PH/21 Added message_body_newlines option.
2448 PH/22 Guard against possible overflow in moan_check_errorcopy().
2450 PH/23 POSIX allows open() to be a macro; guard against that.
2452 PH/24 If the recipient of an error message contained an @ in the local part
2453 (suitably quoted, of course), incorrect values were put in $domain and
2454 $local_part during the evaluation of errors_copy.
2460 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2461 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2464 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2465 issue a MAIL command.
2467 PH/02 In an ACL statement such as
2469 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2471 if a client was not listed at all, or was listed with a value other than
2472 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2473 the condition was not true (as it should be), so access was not denied.
2474 The bug was that the ! inversion was incorrectly passed on to the second
2475 item. This has been fixed.
2477 PH/03 Added additional dnslists conditions == and =& which are different from
2478 = and & when the dns lookup returns more than one IP address.
2480 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2481 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2483 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2484 FSYNC, which compiles an option called disable_fsync that allows for
2485 bypassing fsync(). The documentation is heavily laced with warnings.
2487 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2489 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2490 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2491 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2492 including adding "make clean"; (3) Added -fPIC when compiling the test
2493 dynamically loaded module, to get rid of a warning.
2495 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2496 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2497 = 0s. The bug is otherwise harmless.
2499 PH/07 There was a bug in the dovecot authenticator such that the value of
2500 $auth1 could be overwritten, and so not correctly preserved, after a
2501 successful authentication. This usually meant that the value preserved by
2502 the server_setid option was incorrect.
2504 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2506 PH/09 Installed PCRE release 7.0.
2508 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2509 run for batched SMTP input. It is now run at the start of every message
2510 in the batch. While fixing this I discovered that the process information
2511 (output by running exiwhat) was not always getting set for -bs and -bS
2512 input. This is fixed, and it now also says "batched" for BSMTP.
2514 PH/11 Added control=no_pipelining.
2516 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2517 patch, slightly modified), and move the expansion of helo_data till after
2518 the connection is made in the smtp transport (so it can use these
2521 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2523 PH/14 Added log_selector = +pid.
2525 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2527 PH/16 Add ${if forany and ${if forall.
2529 PH/17 Added dsn_from option to vary the From: line in DSNs.
2531 PH/18 Flush SMTP output before performing a callout, unless control =
2532 no_callout_flush is set.
2534 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2535 was true (the default) a successful delivery failed to delete the retry
2536 item, thus causing premature timeout of the address. The bug is now
2539 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2541 PH/21 Long custom messages for fakedefer and fakereject are now split up
2542 into multiline responses in the same way that messages for "deny" and
2543 other ACL rejections are.
2545 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2546 with slight modification.
2548 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2549 draft, changing the syntax and factoring some duplicate code".
2551 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2552 for deliveries of the second and subsequent messages over the same SMTP
2555 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2556 ${reduce, with only minor "tidies".
2558 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2560 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2561 expansion side effects.
2563 PH/27 When a message times out after an over-quota error from an Exim-imposed
2564 quota, the bounce message says "mailbox is full". This message was not
2565 being given when it was a system quota that was exceeded. It now should
2568 MH/03 Made $recipients available in local_scan(). local_scan() already has
2569 better access to the recipient list through recipients_list[], but
2570 $recipients can be useful in postmaster-provided expansion strings.
2572 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2573 in the case of a MAIL command with additional options following the
2574 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2575 were accidentally chopped off.
2577 PH/29 SMTP synchronization checks are implemented when a command is read -
2578 there is a check that no more input is waiting when there shouldn't be
2579 any. However, for some commands, a delay in an ACL can mean that it is
2580 some time before the response is written. In this time, more input might
2581 arrive, invalidly. So now there are extra checks after an ACL has run for
2582 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2583 pipelining has not been advertised.
2585 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2587 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2588 returned in the SMTP response when smtp_return_error_details was false.
2589 This has been fixed.
2591 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2592 instead of the C I/O that was originally supplied, because problems were
2593 reported on Solaris.
2595 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2596 Exim which did not show up earlier: it was assuming that a call to
2597 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2598 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2599 was a macro that became an assignment, so it seemed to work. This has
2600 changed to a proper function call with a void return, hence the compile
2601 error. Exim's code has been fixed.
2603 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2606 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2608 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2610 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2611 because it turns out to be too big or there is a timeout) there is no
2612 "Completed" line in the log. When some messages of this type were
2613 selected by exigrep, they were listed as "not completed". Others were
2614 picked up by some special patterns. I have improved the selection
2615 criteria to be more general.
2617 PH/38 The host_find_failed option in the manualroute router can now be set
2618 to "ignore", to completely ignore a host whose IP address cannot be
2619 found. If all hosts are ignored, the behaviour is controlled by the new
2620 host_all_ignored option.
2622 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2623 homing or because of multiple MX records with /mx) generated more than
2624 one IP address, and the following item turned out to be the local host,
2625 all the secondary addresses of the first item were incorrectly removed
2626 from the list, along with the local host and any following hosts (which
2627 is what is supposed to happen).
2629 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2630 whoever called Exim into the -H file. In the case of the daemon it was
2631 behaving confusingly. When first started, it used values for whoever
2632 started the daemon, but after a SIGHUP it used the Exim user (because it
2633 calls itself on a restart). I have changed the code so that it now always
2636 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2637 message are rejected with the same error (e.g. no authentication or bad
2638 sender address), and a DATA command is nevertheless sent (as can happen
2639 with PIPELINING or a stupid MUA), the error message that was given to the
2640 RCPT commands is included in the rejection of the DATA command. This is
2641 intended to be helpful for MUAs that show only the final error to their
2644 PH/42 Another patch from the Sieve maintainer.
2646 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2647 Eximstats - Fixed some broken HTML links and added missing column headers
2649 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2650 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2652 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2654 PH/43 Yet another patch from the Sieve maintainer.
2656 PH/44 I found a way to check for a TCP/IP connection going away before sending
2657 the response to the final '.' that terminates a message, but only in the
2658 case where the client has not sent further data following the '.'
2659 (unfortunately, this is allowed). However, in many cases there won't be
2660 any further data because there won't be any more messages to send. A call
2661 to select() can be used: if it shows that the input is "ready", there is
2662 either input waiting, or the socket has been closed. An attempt to read
2663 the next input character can distinguish the two cases. Previously, Exim
2664 would have sent an OK response which the client would never have see.
2665 This could lead to message repetition. This fix should cure that, at
2666 least in a lot of common cases.
2668 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2669 advertised in response to EHLO.
2675 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2676 fixed by 4.65/MH/01 (is this a record?) are fixed:
2678 (i) An empty string was always treated as zero by the numeric comparison
2679 operators. This behaviour has been restored.
2681 (ii) It is documented that the numeric comparison operators always treat
2682 their arguments as decimal numbers. This was broken in that numbers
2683 starting with 0 were being interpreted as octal.
2685 While fixing these problems I realized that there was another issue that
2686 hadn't been noticed. Values of message_size_limit (both the global option
2687 and the transport option) were treated as octal if they started with 0.
2688 The documentation was vague. These values are now always treated as
2689 decimal, and I will make that clear in the documentation.
2695 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2696 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2699 MH/01 Don't check that the operands of numeric comparison operators are
2700 integers when their expansion is in "skipping" mode (fixes bug
2701 introduced by 4.64-PH/07).
2703 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2704 child addresses, Exim now panics and dies. Previously, because the count
2705 is held in a short int, deliveries were likely to be lost. As such a
2706 large number of recipients for a single message is ridiculous
2707 (performance will be very, very poor), I have chosen to impose a limit
2708 rather than extend the field.
2714 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2715 leftover -K file (the existence of which was triggered by #402).
2716 While we were at it, introduced process PID as part of the -K
2717 filename. This should rule out race conditions when creating
2720 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2721 processing considerably. Previous code took too long for large mails,
2722 triggering a timeout which in turn triggers #401.
2724 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2725 in the DK code in transports.c. sendfile() is not really portable,
2726 hence the _LINUX specificness.
2728 TF/01 In the add_headers option to the mail command in an Exim filter,
2729 there was a bug that Exim would claim a syntax error in any
2730 header after the first one which had an odd number of characters
2733 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2734 callout verification, Exim cached a "reject" for the entire domain. This
2735 is correct for most verifications, but it is not correct for a recipient
2736 verification with use_sender or use_postmaster set, because in that case
2737 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2738 case of MAIL FROM:<> rejection from other early rejections (e.g.
2739 rejection of HELO). When verifying a recipient using a non-null MAIL
2740 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2741 Whatever the result of the callout, the value of the domain cache is
2742 left unchanged (for any other kind of callout, getting as far as trying
2743 RCPT means that the domain itself is ok).
2745 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2748 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2749 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2750 session, the daemon ignores ECONNECT errors and logs others; it now
2751 ignores EPIPE as well.
2753 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2754 (quoted-printable decoding).
2756 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2757 later the small subsequent patch to fix an introduced bug.
2759 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2761 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2763 PH/08 An error is now given if message_size_limit is specified negative.
2765 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2766 to be given (somewhat) arbitrary names.
2768 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2771 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2772 miscellaneous code fixes
2774 PH/10 Added the log_reject_target ACL modifier to specify where to log
2777 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2778 hostname. This is wrong, because it relates to the incoming message (and
2779 probably the interface on which it is arriving) and not to the outgoing
2780 callout (which could be using a different interface). This has been
2781 changed to use the value of the helo_data option from the smtp transport
2782 instead - this is what is used when a message is actually being sent. If
2783 there is no remote transport (possible with a router that sets up host
2784 addresses), $smtp_active_hostname is used.
2786 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2787 tweaks were necessary in order to get it to work (see also 21 below):
2788 (a) The code assumed that strncpy() returns a negative number on buffer
2789 overflow, which isn't the case. Replaced with Exim's string_format()
2791 (b) There were several signed/unsigned issues. I just did the minimum
2792 hacking in of casts. There is scope for a larger refactoring.
2793 (c) The code used strcasecmp() which is not a standard C function.
2794 Replaced with Exim's strcmpic() function.
2795 (d) The code set only $1; it now sets $auth1 as well.
2796 (e) A simple test gave the error "authentication client didn't specify
2797 service in request". It would seem that Dovecot has changed its
2798 interface. Fortunately there's a specification; I followed it and
2799 changed what the client sends and it appears to be working now.
2801 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2804 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2805 address A is aliased to B and C, where B exists and C does not. Without
2806 -v the output is "A verified" because verification stops after a
2807 successful redirection if more than one address is generated. However,
2808 with -v the child addresses are also verified. Exim was outputting "A
2809 failed to verify" and then showing the successful verification for C,
2810 with its parentage. It now outputs "B failed to verify", showing B's
2811 parentage before showing the successful verification of C.
2813 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2814 look up a TXT record in a specific list after matching in a combined
2817 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2818 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2819 they consult the DNS. I had assumed they would set it the way they
2820 wanted; and indeed my experiments on Linux seem to show that in some
2821 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2822 To be on the safe side, however, I have now made the interface to
2823 host_find_byname() similar to host_find_bydns(), with an argument
2824 containing the DNS resolver options. The host_find_byname() function now
2825 sets these options at its start, just as host_find_bydns() does. The smtp
2826 transport options dns_qualify_single and dns_search_parents are passed to
2827 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2828 of host_find_byname() use the default settings of RES_DEFNAMES
2829 (qualify_single) but not RES_DNSRCH (search_parents).
2831 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2832 spool_read_header() do less string testing, by means of a preliminary
2833 switch on the second character of optional "-foo" lines. (This is
2834 overdue, caused by the large number of possibilities that now exist.
2835 Originally there were few.) While I was there, I also converted the
2836 str(n)cmp tests so they don't re-test the leading "-" and the first
2837 character, in the hope this might squeeze out yet more improvement.
2839 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2840 flag allowing group syntax was set by the header_syntax check but not
2841 turned off, possible causing trouble later; (2) The flag was not being
2842 set at all for the header_verify test, causing "group"-style headers to
2843 be rejected. I have now set it in this case, and also caused header_
2844 verify to ignore an empty address taken from a group. While doing this, I
2845 came across some other cases where the code for allowing group syntax
2846 while scanning a header line wasn't quite right (mostly, not resetting
2847 the flag correctly in the right place). These bugs could have caused
2848 trouble for malformed header lines. I hope it is now all correct.
2850 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2851 with the "reply" argument non-NULL. The code, however (which originally
2852 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2853 but it didn't always do it. This confused somebody who was copying the
2854 code for some other use. I have removed all the tests.
2856 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2857 feature that was used to support insecure browsers during the U.S. crypto
2858 embargo. It requires special client support, and Exim is probably the
2859 only MTA that supported it -- and would never use it because real RSA is
2860 always available. This code has been removed, because it had the bad
2861 effect of slowing Exim down by computing (never used) parameters for the
2862 RSA_EXPORT functionality.
2864 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2865 authenticator to fail if there's a tab character in the incoming data
2866 (there should never be unless someone is messing about, as it's supposed
2867 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2868 if the connection is using TLS or if the remote IP is the same as the
2869 local IP, and the "valid-client-cert option" if a client certificate has
2872 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2873 authenticators. This can be used for authorization after authentication
2874 succeeds. (In the case of plaintext, it servers for both authentication
2877 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2878 if any retry times were supplied.
2880 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2881 connection, where there is no client IP address to check. In this
2882 situation, the verify now always succeeds.
2884 PH/25 Applied John Jetmore's -Mset patch.
2886 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2888 PH/27 In a string expansion for a processed (not raw) header when multiple
2889 headers of the same name were present, leading whitespace was being
2890 removed from all of them, but trailing whitespace was being removed only
2891 from the last one. Now trailing whitespace is removed from each header
2892 before concatenation. Completely empty headers in a concatenation (as
2893 before) are ignored.
2895 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2896 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2898 PH/29 [Removed. This was a change that I later backed out, and forgot to
2899 correct the ChangeLog entry (that I had efficiently created) before
2900 committing the later change.]
2902 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2903 address errors (4xx response to RCPT) over the same connection as other
2904 messages routed to the same hosts. Such deliveries are always "forced",
2905 so retry times are not inspected. This resulted in far too many retries
2906 for the affected addresses. The effect occurred only when there were more
2907 hosts than the hosts_max_try setting in the smtp transport when it had
2908 the 4xx errors. Those hosts that it had tried were not added to the list
2909 of hosts for which the message was waiting, so if all were tried, there
2910 was no problem. Two fixes have been applied:
2912 (i) If there are any address or message errors in an SMTP delivery, none
2913 of the hosts (tried or untried) are now added to the list of hosts
2914 for which the message is waiting, so the message should not be a
2915 candidate for sending over the same connection that was used for a
2916 successful delivery of some other message. This seems entirely
2917 reasonable: after all the message is NOT "waiting for some host".
2918 This is so "obvious" that I'm not sure why it wasn't done
2919 previously. Hope I haven't missed anything, but it can't do any
2920 harm, as the worst effect is to miss an optimization.
2922 (ii) If, despite (i), such a delivery is accidentally attempted, the
2923 routing retry time is respected, so at least it doesn't keep
2924 hammering the server.
2926 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2927 in ${readsocket because some servers need this prod.
2929 PH/32 Added some extra debug output when updating a wait-xxx database.
2931 PH/33 The hint "could be header name not terminated by colon", which has been
2932 given for certain expansion errors for a long time, was not being given
2933 for the ${if def:h_colon_omitted{... case.
2935 PH/34 The spec says: "With one important exception, whenever a domain list is
2936 being scanned, $domain contains the subject domain." There was at least
2937 one case where this was not true.
2939 PH/35 The error "getsockname() failed: connection reset by peer" was being
2940 written to the panic log as well as the main log, but it isn't really
2941 panic-worthy as it just means the connection died rather early on. I have
2942 removed the panic log writing for the ECONNRESET error when getsockname()
2945 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2946 runs only) independently of the message's sender address. This meant
2947 that, if the 4xx error was in fact related to the sender, a different
2948 message to the same recipient with a different sender could confuse
2949 things. In particular, this can happen when sending to a greylisting
2950 server, but other circumstances could also provoke similar problems.
2951 I have changed the default so that the retry time for these errors is now
2952 based a combination of the sender and recipient addresses. This change
2953 can be overridden by setting address_retry_include_sender=false in the
2956 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2957 remote server are returned as part of bounce messages. This was not
2958 happening for LMTP over a pipe (the lmtp transport), but now it is the
2959 same for both kinds of LMTP.
2961 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2962 in header lines that were in fact CNAMEs. This is no longer the case.
2964 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2965 and queue runs started by the daemon processed all messages. This has
2966 been fixed so that -R and -S can now usefully be given with -q<time>.
2968 PH/40 Import PCRE release 6.7 (fixes some bugs).
2970 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2972 PH/42 Give an error if -q is specified more than once.
2974 PH/43 Renamed the variables $interface_address and $interface_port as
2975 $received_ip_address and $received_port, to make it clear that these
2976 values apply to message reception, and not to the outgoing interface when
2977 a message is delivered. (The old names remain recognized, of course.)
2979 PH/44 There was no timeout on the connect() call when using a Unix domain
2980 socket in the ${readsocket expansion. There now is.
2982 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2983 be meaningful with "accept".
2985 SC/01 Eximstats V1.43
2986 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2988 SC/02 Eximstats V1.44
2989 Use a glob alias rather than an array ref in the generated
2990 parser. This improves both readability and performance.
2992 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2993 Collect SpamAssassin and rejection statistics.
2994 Don't display local sender or destination tables unless
2995 there is data to show.
2996 Added average volumes into the top table text output.
2998 SC/04 Eximstats V1.46
2999 Collect data on the number of addresses (recipients)
3000 as well as the number of messages.
3002 SC/05 Eximstats V1.47
3003 Added 'Message too big' to the list of mail rejection
3004 reasons (thanks to Marco Gaiarin).
3006 SC/06 Eximstats V1.48
3007 Mainlog lines which have GMT offsets and are too short to
3008 have a flag are now skipped.
3010 SC/07 Eximstats V1.49 (Alain Williams)
3011 Added the -emptyok flag.
3013 SC/08 Eximstats V1.50
3014 Fixes for obtaining the IP address from reject messages.
3016 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3017 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3018 whitespace changes from 4.64-PH/27
3020 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3023 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3024 are found, allow negative numbers in numeric criteria)
3026 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3028 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3029 to match changes made in 4.64-PH/43
3031 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3033 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3034 "domain" is actually an IP address, and force a failure. This locks out
3035 those revolvers/nameservers that support "A-for-A" lookups, in
3036 contravention of the specifications.
3038 PH/48 When a host name was looked up from an IP address, and the subsequent
3039 forward lookup of the name timed out, the host name was left in
3040 $sender_host_name, contrary to the specification.
3042 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3043 restricted to single-key lookups, Exim was not diagnosing an error if
3044 * or *@ was used with a query-style lookup.
3046 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3048 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3049 long ago, but noone interested enough thought of it. Let's just say that
3050 the "1.1" means that there are some new functions that weren't there at
3051 some point in the past.
3053 PH/51 Error processing for expansion failure of helo_data from an smtp
3054 transport during callout processing was broken.
3056 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3057 tested/used via the -bh/-bhc/-bs options.
3059 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3060 bug, fixed in subsequent PCRE releases).
3062 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3063 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3065 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3071 SC/01 Use a glob alias rather than an array ref in eximstats generated
3072 parser. This improves both readability and performance.
3074 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3075 Don't display local sender or destination tables in eximstats unless
3076 there is data to show.
3077 Added average volumes into the eximstats top table text output.
3079 SC/03 Collect data on the number of addresses (recipients) as well
3080 as the number of messages in eximstats.
3082 TF/01 Correct an error in the documentation for the redirect router. Exim
3083 does (usually) call initgroups() when daemonizing.
3085 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3086 with consistent privilege compared to when running as a daemon.
3088 TF/03 Note in the spec that $authenticated_id is not set for local
3089 submissions from trusted users.
3091 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3092 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3094 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3095 by adding some example configuration directives to the default
3096 configuration file. A little bit of work is required to uncomment the
3097 directives and define how usernames and passwords are checked, but
3098 there is now a framework to start from.
3100 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3101 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3102 without this. I don't know how relevant this is to other LDAP libraries.
3104 PH/02 Add the verb name to the "unknown ACL verb" error.
3106 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3108 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3110 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3111 directory not expanded when it should be if an expanded home directory
3112 was set for the address (which is overridden by the transport).
3114 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3117 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3118 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3119 because it is too late at that time, and has no effect.
3121 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3122 security issue with \' (bugzilla #107). I could not use the
3123 PQescapeStringConn() function, because it needs a PGconn value as one of
3126 PH/08 When testing addresses using -bt, indicate those final addresses that
3127 are duplicates that would not cause an additional delivery. At least one
3128 person was confused, thinking that -bt output corresponded to deliveries.
3129 (Suppressing duplicates isn't a good idea as you lose the information
3130 about possibly different redirections that led to the duplicates.)
3132 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3133 systems where poll() doesn't work, in particular OS X.
3135 PH/10 Added more information to debugging output for retry time not reached.
3137 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3138 operations in malware.c.
3140 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3143 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3144 syslog_duplication set false, log lines that would normally be written
3145 both the the main log and to the reject log were not written to syslog at
3148 PH/14 In the default configuration, change the use of "message" in ACL warn
3149 statements to "add_header".
3151 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3152 not followed by a command (e.g. "seen endif").
3154 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3155 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3158 PH/17 Added extra conditions to the default value of delay_warning_condition
3162 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3163 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3164 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3167 The Auto-Submitted: and various List- headers are standardised, whereas I
3168 don't think Precedence: ever was.
3170 PH/18 Refactored debugging code in route_finduser() to show more information,
3171 in particular, the error code if getpwnam() issues one.
3173 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3174 This is apparently needed in addition to the PH/07 change above to avoid
3175 any possible encoding problems.
3177 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3178 but not after initializing Perl.
3180 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3181 output them only if debugging. By default they are written stderr,
3182 apparently, which is not desirable.
3184 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3187 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3190 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3192 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3193 authenticated or an ident call has been made. Suppress the default
3194 values for $authenticated_id and $authenticated_sender (but permit -oMai
3195 and -oMas) when testing with -bh.
3197 PH/24 Re-jigged the order of the tests in the default configuration so that the
3198 tests for valid domains and recipients precede the DNS black list and CSA
3199 tests, on the grounds that those ones are more expensive.
3201 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3202 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3203 This bug exists in every version of Exim that I still have, right back to
3206 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3207 However, an attempt to turn on case-sensitivity in a regex key by
3208 including (?-i) didn't work because the subject string was already
3209 lowercased, and the effects were non-intuitive. It turns out that a
3210 one-line patch can be used to allow (?-i) to work as expected.
3216 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3217 other effects) broke the use of negated acl sub-conditions.
3219 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3222 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3223 "Deny" causes Exim to reject the incoming connection with a 554 error.
3224 Unfortunately, if there is a major crisis, such as a disk failure,
3225 tcp-wrappers gives "deny", whereas what one would like would be some
3226 kind of temporary error. A kludge has been added to help with this.
3227 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3228 554 error is used if errno is still zero or contains ENOENT (which occurs
3229 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3232 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3234 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3235 errors. Otherwise a message that provokes a temporary error (when other
3236 messages do not) can cause a whole host to time out.
3238 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3239 addresses were routed directly to files or pipes from a redirect router.
3240 File deliveries just didn't batch; pipe deliveries might have suffered
3243 PH/06 A failure to get a lock for a hints database would erroneously always say
3244 "Failed to get write lock", even when it was really a read lock.
3246 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3247 of 0600. This has been changed to use the value of the lockfile_mode
3248 option (which defaults to 0600).
3250 PH/08 Applied small patch from the Sieve maintainer.
3252 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3253 folder from quota calculations, a direct delivery into this folder messed
3254 up the contents of the maildirsize file. This was because the regex was
3255 used only to exclude .Trash (or whatever) when the size of the mailbox
3256 was calculated. There was no check that a delivery was happening into an
3257 excluded directory. This bug has been fixed by ignoring all quota
3258 processing for deliveries into excluded directories.
3260 PH/10 Added the maildirfolder_create_regex option to appendfile.
3266 PH/01 The code for finding all the local interface addresses on a FreeBSD
3267 system running IPv6 was broken. This may well have applied to all BSD
3268 systems, as well as to others that have similar system calls. The broken
3269 code found IPv4 interfaces correctly, but gave incorrect values for the
3270 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3271 that it would not match correctly against @[] and not recognize the IPv6
3274 PH/02 The ipliteral router was not recognizing addresses of the form user@
3275 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3277 PH/03 Added disable_ipv6.
3279 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3280 decoded form, because it is most often used to construct To: headers
3281 lines in autoreplies, and the decoded form may well be syntactically
3282 invalid. However, $reply_address has leading white space removed, and all
3283 newlines turned into spaces so that the autoreply transport does not
3286 PH/05 If group was specified without a user on a router, and no group or user
3287 was specified on a transport, the group from the router was ignored.
3289 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3290 for visible compile-time settings that can be used to change these
3291 numbers, for those that want even more. Backwards compatibility with old
3292 spool files has been maintained. However, going back to a previous Exim
3293 release will lost any variables that are in spool files.
3295 PH/07 Two small changes when running in the test harness: increase delay when
3296 passing a TCP/IP connection to a new process, in case the original
3297 process has to generate a bounce, and remove special handling of
3298 127.0.0.2 (sic), which is no longer necessary.
3300 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3301 be the same on different OS.
3303 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3306 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3307 whether --show-vars was specified or not
3309 JJ/02 exipick: Added support for new ACL variable spool format introduced
3312 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3313 syntactically invalid From: or Reply-to: line, and a filter used this to
3314 generate an autoreply, and therefore failed to obtain an address for the
3315 autoreply, Exim could try to deliver to a non-existent relative file
3316 name, causing unrelated and misleading errors. What now happens is that
3317 it logs this as a hard delivery error, but does not attempt to create a
3320 PH/11 The exinext utility has a -C option for testing purposes, but although
3321 the given file was scanned by exinext itself; it wasn't being passed on
3322 when Exim was called.
3324 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3325 an end-of-file indication when reading a command response.
3327 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3328 compiled. In many other places in Exim, IPv6 addresses are always
3329 recognized, so I have changed this. It also means that IPv4 domain
3330 literals of the form [IPV4:n.n.n.n] are now always recognized.
3332 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3333 used if the router is not running as root, for example, when verifying at
3334 ACL time, or when using -bh. The debugging output from this situation was
3335 non-existent - all you got was a failure to exec. I have made two
3338 (a) Failures to set uid/gid, the current directory, or a process leader
3339 in a subprocess such as that created by queryprogram now generate
3340 suitable debugging output when -d is set.
3342 (b) The queryprogram router detects when it is not running as root,
3343 outputs suitable debugging information if -d is set, and then runs
3344 the subprocess without attempting to change uid/gid.
3346 PH/15 Minor change to Makefile for building test_host (undocumented testing
3349 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3350 additional section of a DNS packet that returns MX or SRV records.
3351 Instead, it always explicitly searches for A/AAAA records. This avoids
3352 major problems that occur when a DNS server includes only records of one
3353 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3354 fixed another bug: if SRV records were looked up and the corresponding
3355 address records were *not* found in the additional section, the port
3356 values from the SRV records were lost.
3358 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3359 using the correct key (the original address) when searching the retry
3360 rules in order to find which one to use for generating the retry hint.
3362 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3363 adding the default one. Similarly, if it contains a Reply-To: header, the
3364 errors_reply_to option, if set, is not used.
3366 PH/19 When calculating a retry time, Exim used to measure the "time since
3367 failure" by looking at the "first failed" field in the retry record. Now
3368 it does not use this if it is later than than the arrival time of the
3369 message. Instead it uses the arrival time. This makes for better
3370 behaviour in cases where some deliveries succeed, thus re-setting the
3371 "first failed" field. An example is a quota failure for a huge message
3372 when small messages continue to be delivered. Without this change, the
3373 "time since failure" will always be short, possible causing more frequent
3374 delivery attempts for the huge message than are intended.
3375 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3377 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3378 $1, $2, $3) because the numerical variables can be reset during some
3379 expansion items (e.g. "match"), thereby losing the authentication data.
3381 PH/21 Make -bV show the size of off_t variables so that the test suite can
3382 decide whether to run tests for quotas > 2G.
3384 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3385 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3386 filecount value is greater than 2G or if a quota value is greater than 2G
3387 on a system where the size of off_t is not greater than 4, a panic error
3390 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3391 never match. The debug and -bh output now contains an explicit error
3392 message indicating a malformed IPv4 address or mask.
3394 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3395 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3396 PH/23 above applies.
3398 PH/25 Do not write to syslog when running in the test harness. The only
3399 occasion when this arises is a failure to open the main or panic logs
3400 (for which there is an explicit test).
3402 PH/26 Added the /no_tell option to "control=freeze".
3404 PH/27 If a host name lookup failed very early in a connection, for example, if
3405 the IP address matched host_lookup and the reverse lookup yielded a name
3406 that did not have a forward lookup, an error message of the form "no IP
3407 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3408 could be logged. Now it outputs the IP address instead of "NULL".
3410 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3411 allows the sender and the authenticated sender to be set when
3412 submitting a message from within Exim. Since child_open_exim() is
3413 documented for local_scan(), the new function should be too.
3415 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3416 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3417 results in an empty string is now treated as unset.
3419 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3421 PH/31 Added sender_verify_fail logging option.
3423 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3424 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3425 tidied the source and removed it altogether.
3427 PH/33 When a queue run was abandoned because the load average was too high, a
3428 log line was always written; now it is written only if the queue_run log
3429 selector is set. In addition, the log line for abandonment now contains
3430 information about the queue run such as the pid. This is always present
3431 in "start" and "stop" lines but was omitted from the "abandon" line.
3433 PH/34 Omit spaces between a header name and the colon in the error message that
3434 is given when verify = headers_syntax fails (if there are lots of them,
3435 the message gets confusing).
3437 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3438 names, as there are now some PTR records that contain slashes. This check
3439 is only to protect against broken name servers that fall over on strange
3440 characters, so the fact that it applies to all lookups doesn't matter.
3442 PH/36 Now that the new test suite is complete, we can remove some of the
3443 special code in Exim that was needed for the old test suite. For example,
3444 sorting DNS records because real resolvers return them in an arbitrary
3445 order. The new test suite's fake resolver always returns records in the
3448 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3449 bounces) except when queue_only is set, to avoid logging races between
3450 the different processes.
3452 PH/38 Panic-die if .include specifies a non-absolute path.
3454 PH/39 A tweak to the "H" retry rule from its user.
3456 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3457 a label. They prevented compilation on older perls.
3459 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3460 a warning to be raised on newish perls.
3462 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3463 on queue. Changes to match documented behaviour of showing count of
3464 messages matching specified criteria.
3466 PH/40 Changed the default ident timeout from 30s to 5s.
3468 PH/41 Added support for the use of login_cap features, on those BSD systems
3469 that have them, for controlling the resources used by pipe deliveries.
3471 PH/42 The content-scanning code uses fopen() to create files in which to put
3472 message data. Previously it was not paying any attention to the mode of
3473 the files. Exim runs with umask(0) because the rest of the code creates
3474 files with open(), and sets the required mode explicitly. Thus, these
3475 files were ending up world-writeable. This was not a big issue, because,
3476 being within the spool directory, they were not world-accessible. I have
3477 created a function called modefopen, which takes an additional mode
3478 argument. It sets umask(777), creates the file, chmods it to the required
3479 mode, then resets the umask. All the relevant calls to fopen() in the
3480 content scanning code have been changed to use this function.
3482 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3483 to 24 hours. This avoids potential overflow problems when processing G
3484 and H retry rules. I suspect nobody ever tinkers with this value.
3486 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3488 PH/45 When the plaintext authenticator is running as a client, the server's
3489 challenges are checked to ensure they are valid base64 strings. By
3490 default, the authentication attempt is cancelled if an invalid string is
3491 received. Setting client_ignore_invalid_base64 true ignores these errors.
3492 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3493 they are received. Thus, the responses can be made to depend on the
3494 challenges. If an invalid string is ignored, an empty string is placed in
3497 PH/46 Messages that are created by the autoreply transport now contains a
3498 References: header, in accordance with RFCs 2822 and 3834.
3500 PH/47 Added authenticated_sender_force to the smtp transport.
3502 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3504 PH/49 Installed latest patch from the Sieve maintainer.
3506 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3507 was also set, the appendfile transport was unnecessarily scanning a
3508 directory of message files (e.g. for maildir delivery) to find the count
3509 of files (along with the size), even though it did not need this
3510 information. It now does the scan only if it needs to find either the
3511 size of the count of files.
3513 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3515 PH/52 Two bugs concerned with error handling when the smtp transport is
3518 (i) Exim was not creating retry information for temporary errors given
3519 for individual recipients after the DATA command when the smtp transport
3520 was used in LMTP mode. This meant that they could be retried too
3521 frequently, and not timed out correctly.
3523 (ii) Exim was setting the flag that allows error details to be returned
3524 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3525 recipients that were returned after the DATA command.
3527 PH/53 This is related to PH/52, but is more general: for any failing address,
3528 when detailed error information was permitted to be returned to the
3529 sender, but the error was temporary, then after the final timeout, only
3530 "retry timeout exceeded" was returned. Now it returns the full error as
3531 well as "retry timeout exceeded".
3533 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3534 do this, and (what is worse) MTAs that accept it.
3536 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3537 will now be deprecated.
3539 PH/56 New os.c-cygwin from the Cygwin maintainer.
3541 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3542 formats (previously only available in exim formats via -bpr, -bpru,
3543 and -bpra. Now also available in native and exiqgrep formats)
3545 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3546 with very large, slow to parse queues
3548 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3550 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3552 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3553 responses to authentication challenges, though it was showing the
3554 challenges; (ii) I've removed the CR characters from the debug output for
3557 PH/58 Allow for the insertion of a newline as well as a space when a string
3558 is turned into more than one encoded-word during RFC 2047 encoding. The
3559 Sieve code now uses this.
3561 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3562 data_4xx, lost_connection, tls_required.
3564 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3565 message was being sent as an SMTP response.
3567 PH/61 Add -l and -k options to exicyclog.
3569 PH/62 When verifying, if an address was redirected to one new address, so that
3570 verification continued, and the new address failed or deferred after
3571 having set something in $address_data, the value of $address_data was not
3572 passed back to the ACL. This was different to the case when no
3573 redirection occurred. The value is now passed back in both cases.
3575 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3576 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3577 use login_cap.h, so on its own it isn't the distinguishing feature. The
3578 new name refers directly to the setclassresources() function.
3580 PH/65 Added configuration files for NetBSD3.
3582 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3584 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3585 is preferred over IPv4.
3587 PH/68 The bounce_return_message and bounce_return_body options were not being
3588 honoured for bounces generated during the reception of non-SMTP messages.
3589 In particular, this applied to messages rejected by the ACL. This bug has
3590 been fixed. However, if bounce_return_message is true and bounce_return_
3591 body is false, the headers that are returned for a non-SMTP message
3592 include only those that have been read before the error was detected.
3593 (In the case of an ACL rejection, they have all been read.)
3595 PH/69 The HTML version of the specification is now built in a directory called
3596 spec_html instead of spec.html, because the latter looks like a path with
3597 a MIME-type, and this confuses some software.
3599 PH/70 Catch two compiler warnings in sieve.c.
3601 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3602 function verify_get_ident() calls ip_connect() to connect a socket, but
3603 if the "connect()" function timed out, ip_connect() used to close the
3604 socket. However, verify_get_ident() also closes the socket later, and in
3605 between Exim writes to the log, which may get opened at this point. When
3606 the socket was closed in ip_connect(), the log could get the same file
3607 descriptor number as the socket. This naturally causes chaos. The fix is
3608 not to close the socket in ip_connect(); the socket should be closed by
3609 the function that creates it. There was only one place in the code where
3610 this was missing, in the iplookup router, which I don't think anybody now
3611 uses, but I've fixed it anyway.
3613 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3614 well as to direct DNS lookups. Otherwise the handling of names in host
3615 lists is inconsistent and therefore confusing.
3621 PH/01 Two changes to the default runtime configuration:
3623 (1) Move the checks for relay_from_hosts and authenticated clients from
3624 after to before the (commented out) DNS black list checks.
3626 (2) Add control=submission to the relay_from_hosts and authenticated
3627 clients checks, on the grounds that messages accepted by these
3628 statements are most likely to be submissions.
3630 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3632 (1) Generate an error if the third argument for the ${prvs expansion is
3635 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3638 (3) Reset the variables that are obtained from the first argument of
3639 ${prvscheck and used in the second argument before leaving the code,
3640 because their memory is reclaimed, so using them afterwards may do
3643 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3644 one (it's much easier than Tom thought :-).
3646 (5) Because of (4), we can now allow for the use of $prvscheck_result
3647 inside the third argument.
3649 PH/03 For some reason, the default setting of PATH when running a command from
3650 a pipe transport was just "/usr/bin". I have changed it to
3653 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3654 anything to be listed in the output from -bV.
3656 PH/05 When a filter generated an autoreply, the entire To: header line was
3657 quoted in the delivery log line, like this:
3659 => >A.N.Other <ano@some.domain> <original@ddress> ...
3661 This has been changed so that it extracts the operative address. There
3662 may be more than one such address. If so, they are comma-separated, like
3665 => >ano@some.domain,ona@other.domain <original@ddress> ...
3667 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3668 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3669 not being looked up in the rDNS to get a host name, Exim was showing the
3670 IP address twice in Received: lines, even though the IP addresses were
3671 identical. For example:
3673 Received: from [1.2.3.4] (helo=[1.2.3.4])
3675 However, if the real host name was known, it was omitting the HELO data
3676 if it matched the actual IP address. This has been tidied up so that it
3677 doesn't show the same IP address twice.
3679 PH/07 When both +timestamp and +memory debugging was on, the value given by
3680 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3681 called by the debug printing, thereby overwriting the timestamp buffer.
3682 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3684 PH/08 When the original message was included in an autoreply transport, it
3685 always said "this is a copy of the message, including all the headers",
3686 even if body_only or headers_only was set. It now gives an appropriate
3689 PH/09 Applied a patch from the Sieve maintainer which:
3691 o fixes some comments
3692 o adds the (disabled) notify extension core
3693 o adds some debug output for the result of if/elsif tests
3694 o points to the current vacation draft in the documentation
3695 and documents the missing references header update
3699 o fixes a bug in processing the envelope test (when testing
3700 multiple envelope elements, the last element determined the
3703 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3704 Electronic Mail") by including:
3706 Auto-submitted: auto-generated
3708 in the messages that it generates (bounce messages and others, such as
3709 warnings). In the case of bounce messages for non-SMTP messages, there was
3710 also a typo: it was using "Auto_submitted" (underscore instead of
3711 hyphen). Since every message generated by Exim is necessarily in response
3712 to another message, thes have all been changed to:
3714 Auto-Submitted: auto-replied
3716 in accordance with these statements in the RFC:
3718 The auto-replied keyword:
3720 - SHOULD be used on messages sent in direct response to another
3721 message by an automatic process,
3723 - MUST NOT be used on manually-generated messages,
3725 - MAY be used on Delivery Status Notifications (DSNs) and Message
3726 Disposition Notifications (MDNs),
3728 - MUST NOT be used on messages generated by automatic or periodic
3729 processes, except for messages which are automatic responses to
3732 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3733 to the default Received: header definition.
3735 PH/12 Added log selector acl_warn_skipped (default on).
3737 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3738 variables because (a) they are in the wrong storage pool and (b) even if
3739 they were copied, it wouldn't work properly because of the caching.
3741 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3742 checking when decoding. Apparently there are clients that generate
3743 overlong encoded strings. Why am I not surprised?
3745 PH/15 If the first argument of "${if match_address" was not empty, but did not
3746 contain an "@" character, Exim crashed. Now it writes a panic log message
3747 and treats the condition as false.
3749 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3751 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3752 extension "envelope-auth". The code is finished and in agreement with
3753 other implementations, but there is no documentation so far and in fact,
3754 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3755 not changing the active code.
3757 Print executed "if" and "elsif" statements when debugging is used. This
3758 helps a great deal to understand what a filter does.
3760 Document more things not specified clearly in RFC3028. I had all this
3761 sorted out, when out of a sudden new issues came to my mind. Oops."
3763 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3766 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3767 canonical form (as documented). However, after a host name lookup from
3768 the IP address, check_host() was doing a simple string comparison with
3769 addresses acquired from the DNS when checking that the found name did
3770 have the original IP as one of its addresses. Since any found IPv6
3771 addresses are likely to be in abbreviated form, the comparison could
3772 fail. Luckily, there already exists a function for doing the comparison
3773 by converting both addresses to binary, so now that is used instead of
3774 the text comparison.
3776 PH/20 There was another similar case to PH/19, when a complete host name was
3777 given in a host list; looking up its IP address could give an abbreviated
3778 form, whereas the current host's name might or might not be abbreviated.
3779 The same fix has been applied.
3785 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3786 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3789 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3790 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3792 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3794 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3795 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3796 Also, for "Auto-Submitted", treat anything other than "no" as
3797 non-personal, in accordance with RFC 3834. (Previously it treated
3798 anything starting "auto-" as non-personal.)
3800 TF/01 The control=submission/name=... option had a problem with syntax
3801 errors if the name included a slash character. The /name= option
3802 now slurps the rest of the string, so it can include any characters
3803 but it must come last in the list of options (after /sender_retain
3806 PH/05 Some modifications to the interface to the fake nameserver for the new
3814 TK/01 Added the "success_on_redirect" address verification option. See
3815 NewStuff for rationale and an example.
3817 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3819 PH/02 Patch to exigrep to allow it to work on syslog lines.
3821 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3822 fread() to skip over the body file's header line, because in Cygwin the
3823 header line is locked and is inaccessible.
3825 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3826 co-exist for some time) to make it clear that it is the Exim ID that is
3827 referenced, not the Message-ID: header line.
3829 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3830 string_format() function, because snprintf() does not exist on all
3833 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3834 ${stat: expansion item.
3836 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3837 protocol synchronization error", to keep the pedants happy.
3839 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3840 well as for IRIX systems, when gcc is being used. See the host.c source
3843 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3845 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3848 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3849 transport and to the smtp transport in LMTP mode.
3851 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3853 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3854 run a filter in a subprocess. This could lead to confusion in subsequent
3855 lookups in the parent process. There should also be a search_tidyup() at
3856 the end of the subprocess.
3858 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3859 only if the host matched helo_try_verify_hosts, which caused the
3860 verification to occur when the EHLO/HELO command was issued. The ACL just
3861 tested the remembered result. Now, if a previous verification attempt has
3862 not happened, "verify = helo" does it there and then.
3864 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3866 TK/03 Fix log output including CR from clamd.
3868 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3869 exist provoked a memory error which could cause a segfault.
3871 PH/15 Installed PCRE 6.2
3873 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3875 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3876 of the problem. Specifically, suggested +O2 rather than +O1 for the
3879 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3881 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3882 record, the port was not correctly logged when the outgoing_port log
3883 selector was set (it logged the transort's default port).
3885 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3886 fallback_hosts, and "hosts" in the smtp transport.
3888 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3889 host errors such as "Connection refused".
3891 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3892 authentication with radiusclient 0.4.9:
3894 - Error returned from rc_read_config was caught wrongly
3895 - Username/password not passed on to radius server due to wrong length.
3897 The presumption is that some radiusclient API changes for 4.51/PH/17
3898 were not taken care of correctly. The code is still untested by me (my
3899 Linux distribution still has 0.3.2 of radiusclient), but it was
3900 contributed by a Radius user.
3902 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3903 expanding the "port" option of the smtp transport.
3905 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3906 while reading a MIME header. Thanks to Tom Hughes for a patch.
3908 PH/24 Include config.h inside local_scan.h so that configuration settings are
3911 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3912 that in an ACL for RCPT (for example), you can examine exactly what was
3915 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3916 commands, but it was not correctly comparing the address with the actual
3917 client host address. Thus, it would show the EHLO address in Received:
3918 header lines when this was not necessary.
3920 PH/27 Added the % operator to ${eval:}.
3922 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3923 it should be ignoring failures (because with -C, for example, it has lost
3924 privilege). It wasn't ignoring creation failures other than "already
3927 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3930 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3931 because an input line was too long, either on its own, or by virtue of
3932 too many continuations, the temporary file was not being removed, and the
3933 return code was incorrect.
3935 PH/31 Missing "BOOL" in function definition in filtertest.c.
3937 PH/32 Applied Sieve patches from the maintainer.
3939 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3941 PH/33 Added "verify = not_blind".
3943 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3944 Local/Makefile (with some defaults set). These are used in built scripts
3945 such as exicyclog, but they have never been used in the exim_install
3946 script (though there are many overriding facilities there). I have
3947 arranged that the exim_install script now takes note of these two
3950 PH/35 Installed configuration files for Dragonfly.
3952 PH/36 When a locally submitted message by a trusted user did not contain a
3953 From: header, and the sender address was obtained from -f or from an SMTP
3954 MAIL command, and the trusted user did not use -F to supply a sender
3955 name, $originator_name was incorrectly used when constructing a From:
3956 header. Furthermore, $originator_name was used for submission mode
3957 messages from external hosts without From: headers in a similar way,
3958 which is clearly wrong.
3960 PH/37 Added control=suppress_local_fixups.
3962 PH/38 When log_selector = +received_sender was set, and the addition of the
3963 sender made the log line's construction buffer exactly full, or one byte
3964 less than full, an overflow happened when the terminating "\n" was
3967 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3968 when the result of a list match is failure because a DNS lookup failed.
3970 PH/40 RM_COMMAND is now used in the building process.
3972 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3973 the "build-* directories that it finds.
3975 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3976 address in a domain literal was a prefix of an interface address.
3978 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3979 when verifying a sender address, unless rewrite_headers is false.
3981 PH/44 Wrote a long comment about why errors_to addresses are verified as
3982 recipients, not senders.
3984 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3985 the ratelimit ACL was added.
3987 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3989 PH/46 Added extra information about PostgreSQL errors to the error string.
3991 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3992 avoiding the need to install special zones in a real server. This is
3993 backwards compatible; if it can't find the fake resolver, it drops back.
3994 Thus, both old and new test suites can be run.
3996 TF/02 Added util/ratelimit.pl
3998 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3999 clock is set back in time.
4001 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4002 Candler <B.Candler@pobox.com>.
4004 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4005 OK for addresses that are the result of redirections.
4007 PH/48 A number of further additions for the benefit of the new test suite,
4008 including a fake gethostbyname() that interfaces to the fake DNS resolver
4011 TF/06 The fix for widen_domains has also been applied to qualify_single and
4012 search_parents which are the other dnslookup options that can cause
4015 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4018 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4020 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4021 service" if it encountered an MX record with an empty target hostname.
4022 The message is now "an MX or SRV record indicated no SMTP service".
4024 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4025 if the DNS of the sending site is misconfigured. This is quite a
4026 common situation. This change restores the behaviour of treating a
4027 helo verification defer as a failure.
4029 PH/51 If self=fail was set on a router, the bounce message did not include the
4030 actual error message.
4036 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4038 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4039 command itself ended in error, the underlying message about the transport
4040 filter timeout was being overwritten with the pipe command error. Now the
4041 underlying error message should be appended to the second error message.
4043 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4045 PH/02 Reduce the amount of output that "make" produces by default. Full output
4046 can still be requested.
4048 PH/03 The warning log line about a condition test deferring for a "warn" verb
4049 was being output only once per connection, rather than after each
4050 occurrence (because it was using the same function as for successful
4051 "warn" verbs). This seems wrong, so I have changed it.
4053 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4054 it should not have, which might have caused a crash in the right
4055 circumstances, but probably never did.
4057 PH/04 Installed a modified version of Tony Finch's patch to make submission
4058 mode fix the return path as well as the Sender: header line, and to
4059 add a /name= option so that you can make the user's friendly name appear
4062 TF/03 Added the control = fakedefer ACL modifier.
4064 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4065 Mark Lowes for thorough testing.
4067 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4069 TK/03 Merged latest SRS patch from Miles Wilton.
4071 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4072 with the definition in sysexits.h (which is #included earlier).
4073 Fortunately, Exim does not actually use EX_OK. The code used to try to
4074 preserve the sysexits.h value, by assuming that macro definitions were
4075 scanned for macro replacements. I have been disabused of this notion,
4076 so now the code just undefines EX_OK before #including unistd.h.
4078 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4079 in the smtp transport. When a block could not be written in a single
4080 write() function, the timeout was being re-applied to each part-write.
4081 This seems wrong - if the receiver was accepting one byte at a time it
4082 would take for ever. The timeout is now adjusted when this happens. It
4083 doesn't have to be particularly precise.
4085 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4086 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4088 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4089 without a domain if the check to <postmaster@domain> fails.
4091 SC/01 Eximstats: added -xls and the ability to specify output files
4092 (patch written by Frank Heydlauf).
4094 SC/02 Eximstats: use FileHandles for outputting results.
4096 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4098 SC/04 Eximstats: fixed display of large numbers with -nvr option
4100 SC/05 Eximstats: fixed merging of reports with empty tables.
4102 SC/06 Eximstats: added the -include_original_destination flag
4104 SC/07 Eximstats: removed tabs and trailing whitespace.
4106 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4108 TK/06 MBOX spool code: Add real "From " MBOX separator line
4109 so the .eml file is really in mbox format (even though
4110 most programs do not really care). Patch from Alex Miller.
4112 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4113 The latter is generated from $received_to and is only set if the
4114 message has one envelope recipient. SA can use these headers,
4115 obviously out-of-the-box. Patch from Alex Miller.
4117 PH/08 The ${def test on a variable was returning false if the variable's
4118 value was "0", contrary to what the specification has always said!
4119 The result should be true unless the variable is empty.
4121 PH/09 The syntax error of a character other than { following "${if
4122 def:variable_name" (after optional whitespace) was not being diagnosed.
4123 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4124 accidental colon was present, for example, could give incorrect results.
4126 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4127 result is used (not including appendfile, where other changes are about
4130 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4131 This involved changing a lot of size variables from int to off_t. It
4132 should work with maildirs and everything.
4134 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4135 spamd dying while we are connected to it.
4137 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4140 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4141 function for BDB 4.3.
4143 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4145 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4146 the sources for PCRE have been re-arranged and more files are now
4149 PH/15 The code I had for printing potentially long long variables in PH/11
4150 above was not the best (it lost precision). The length of off_t variables
4151 is now inspected at build time, and an appropriate printing format (%ld
4152 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4153 to be "long long int" or "long int". This is needed for the internal
4154 formatting function string_vformat().
4156 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4157 the configuration file to be ":syslog", then the script "guesses" where
4158 the logs files are, rather than using the compiled in default. In our
4159 case the guess is not the same as the compiled default, so the script
4160 suddenly stopped working when I started to use syslog. The patch checks
4161 to see if log_file_path is "". If so, it attempts to read it from exim
4162 with no configuration file to get the compiled in version, before it
4163 falls back to the previous guessing code."
4165 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4166 implementing BATV in an Exim configuration. See NewStuff for the gory
4169 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4170 Makefile that are specific to HP-UX.
4172 PH/18 If the "use_postmaster" option was set for a recipient callout together
4173 with the "random" option, the postmaster address was used as the MAIL
4174 FROM address for the random test, but not for the subsequent recipient
4175 test. It is now used for both.
4177 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4178 patch removes a few documentation additions to RFC 3028, because the
4179 latest draft now contains them. It adds the new en;ascii-case comparator
4180 and a new error check for 8bit text in MIME parts. Comparator and
4181 require names are now matched exactly. I enabled the subaddress
4182 extension, but it is not well tested yet (read: it works for me)."
4184 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4185 rework some of the code of TK/09 above to avoid the hardwired use of
4186 "%lld" and "long long". Replaced the call to snprintf() with a call to
4189 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4190 records point to non-existent hosts", "retry timeout exceeded", and
4191 "retry time not reached for any host after a long failure period".
4193 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4194 experimental DomainKeys support:
4196 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4197 (2) On an error such as an illegally used "control", the wrong name for
4198 the control was given.
4200 These problems did NOT occur unless DomainKeys support was compiled.
4202 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4204 PH/24 Added ${if match_ip condition.
4206 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4207 databases so that it will be absolutely obvious if a crash occurs in the
4208 DB library. This is a regular occurrence (often caused by mis-matched
4211 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4212 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4213 picked up on a user's system that detects such things. There doesn't seem
4214 to be a gcc warning option for this - only an attribute that has to be
4215 put on the function's prototype. It seems that in Fedora Core 4 they have
4216 set this on a number of new functions. No doubt there will be more in due
4219 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4220 specify a transport. However, if an address that was verified by such a
4221 router was the subject of a callout, Exim crashed because it tried to
4222 read the rcpt_include_affixes from the non-existent transport. Now it
4223 just assumes that the setting of that option is false. This bug was
4224 introduced by 4.51/PH/31.
4226 PH/28 Changed -d+all to exclude +memory, because that information is very
4227 rarely of interest, but it makes the output a lot bigger. People tend to
4228 do -d+all out of habit.
4230 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4231 code in os-type was giving problems when libc.so lives in lib64, like on
4234 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4235 aren't the modern standard, and it seems that some systems' include files
4236 don't always have them. Exim was already checking for some of the newer
4237 ones like T_AAAA, and defining it itself. I've added checks for all the
4238 record types that Exim uses.
4240 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4241 not automatically generating a new one, as it is supposed to. This
4242 prevented TLS from working. If the file did exist, but contained invalid
4243 data, a new version was generated, as expected. It was only the case of a
4244 non-existent file that was broken.
4246 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4247 with a change in libdomainkeys > 0.64.
4249 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4250 from DNS. If the selector record carries the flag, it now has
4251 precedence over the domain-wide flag.
4253 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4255 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4256 the use of a port name that isn't defined in /etc/services) occurred, the
4257 message was deferred as in a normal delivery, and thus remained on the
4258 spool, instead of being failed because of the mua_wrapper setting. This
4259 is now fixed, and I tidied up some of the mua_wrapper messages at the
4262 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4263 the messages in a hash of arrays rather than using individual hashes.
4264 This is a bit cleaner and results in dramatic memory savings, albeit
4265 at a slight CPU cost.
4267 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4268 as requested by Marc Sherman.
4270 SC/10 Eximstats: added histograms for user specified patterns as requested
4273 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4275 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4276 fopen() in the content-scanning modules that did not already have it.
4282 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4283 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4285 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4287 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4289 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4290 to test Sieve filters that use "vacation".
4292 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4293 that changes the way the GnuTLS parameters are stored in the cache file.
4294 The new format can be generated externally. For backward compatibility,
4295 if the data in the cache doesn't make sense, Exim assumes it has read an
4296 old-format file, and it generates new data and writes a new file. This
4297 means that you can't go back to an older release without removing the
4300 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4301 work if there are any delivery delays because "one_time" forces the
4302 parent to be marked "delivered", so its unseen clone is never tried
4303 again. For this reason, Exim now forbids the simultaneous setting of
4306 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4307 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4308 entry if you want to know the details. The fix, however, neglected to
4309 consider the case when local delivery batching is involved. The test for
4310 "previously delivered" was not happening when checking to see if an
4311 address could be batched with a previous (undelivered) one; under
4312 certain circumstances this could lead to multiple deliveries to the same
4315 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4316 in its include files, and this causes problems building Exim.
4318 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4319 header_syntax) but Exim was just ignoring anything given after a slash.
4320 In particular, this caused confusion with an attempt to use "verify =
4321 reverse_host_lookup/defer_ok". An error is now given when options are
4322 supplied for verify items that do not have them. (Maybe reverse_host_
4323 lookup should have a defer_ok option, but that's a different point.)
4325 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4326 defined by RFC 821) to 2048, because there were problems with some AUTH
4327 commands, and RFC 1869 says the size should be increased for extended
4328 SMTP commands that take arguments.
4330 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4333 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4334 "unknown" error; now it says that the functionality isn't in the binary.
4336 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4337 an address' error message when a string expansion fails (syntax or
4338 whatever). Otherwise the password may appear in the log. Following change
4339 PH/42 below, there is no longer a chance of it appearing in a bounce
4342 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4344 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4345 of its addresses was ever tried. (Bugzilla bug #2.)
4347 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4348 the result incorrectly in the debug output. (It correctly added a newline
4349 to what was transported.)
4351 TF/01 Added $received_time.
4353 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4354 commented out examples of how to interface to a virus scanner and to
4355 SpamAssassin. Also added commented examples of av_scanner and
4356 spamd_address settings.
4358 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4359 and controls are allowed in which ACLs. There were a couple of minor
4360 errors. Some of the entries in the conditions table (which is a table of
4361 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4362 negation of where the condition IS allowed.
4364 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4366 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4367 header file does not have a version number, so I've had to invent a new
4368 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4369 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4370 radiusclient), but it was contributed by a Radius user.
4372 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4373 files or directories, for OpenSSL.
4375 PH/19 When an Exim process that is running as root has to create an Exim log
4376 file, it does so in a subprocess that runs as exim:exim so as to get the
4377 ownership right at creation (otherwise, other Exim processes might see
4378 the file with the wrong ownership). There was no test for failure of this
4379 fork() call, which would lead to the process getting stuck as it waited
4380 for a non-existent subprocess. Forks do occasionally fail when resources
4381 run out. I reviewed all the other calls to fork(); they all seem to check
4384 PH/20 When checking for unexpected SMTP input at connect time (before writing
4385 the banner), Exim was not dealing correctly with a non-positive return
4386 from the read() function. If the client had disconnected by this time,
4387 the result was a log entry for a synchronization error with an empty
4388 string after "input=" when read() returned zero. If read() returned -1
4389 (an event I could not check), uninitialized data bytes were printed.
4390 There were reports of junk text (parts of files, etc) appearing after
4393 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4395 PH/22 Added support for macro redefinition, and (re)definition in between
4396 driver and ACL definitions.
4398 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4399 forgetting to use the resulting value; it was using the unexpanded value.
4401 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4402 hadn't been configured. The fix is from Juergen Kreileder, who
4403 understands it better than I do:
4405 "Here's what I see happening with three configured cyrus_sasl
4406 authenticators configured (plain, login, cram-md5):
4408 On startup auth_cyrus_sasl_init() gets called for each of these.
4409 This means three calls to sasl_listmech() without a specified mech_list.
4410 => SASL tests which mechs of all available mechs actually work
4411 => three warnings about OTP not working
4412 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4414 With the patch, sasl_listmech() also gets called three times. But now
4415 SASL's mech_list option is set to the server_mech specified in the the
4416 authenticator. Or in other words, the answer from sasl_listmech()
4417 gets limited to just the mech you're testing for (which is different
4419 => the return list contains just 'plain' or 'login', 'cram-md5' or
4420 nothing depending on the value of ob->server_mech.
4422 I've just tested the patch: Authentication still works fine,
4423 unavailable mechs specified in the exim configuration are still
4424 caught, and the auth.log warnings about OTP are gone."
4426 PH/25 When debugging is enabled, the contents of the command line are added
4427 to the debugging output, even when log_selector=+arguments is not
4430 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4431 answer is "GNU", and only if the return is "GNU/something" is the answer
4434 PH/27 $acl_verify_message is now set immediately after the failure of a
4435 verification in an ACL, and so is available in subsequent modifiers. In
4436 particular, the message can be preserved by coding like this:
4438 warn !verify = sender
4439 set acl_m0 = $acl_verify_message
4441 Previously, $acl_verify_message was set only while expanding "message"
4442 and "log_message" when a very denied access.
4444 PH/28 Modified OS/os.c-Linux with
4446 -#ifndef OS_LOAD_AVERAGE
4447 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4449 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4450 nomenclature these days.)
4452 PH/29 Installed patch from the Sieve maintainer that adds the options
4453 sieve_useraddress and sieve_subaddress to the redirect router.
4455 PH/30 In these circumstances:
4456 . Two addresses routed to the same list of hosts;
4457 . First host does not offer TLS;
4458 . First host accepts first address;
4459 . First host gives temporary error to second address;
4460 . Second host offers TLS and a TLS session is established;
4461 . Second host accepts second address.
4462 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4463 and peerdn, if requested) that were in fact used only for the second
4466 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4467 attention to any local part prefix or suffix that was matched by the
4468 router that accepted the address. It now behaves in the same way as it
4469 does for delivery: the affixes are removed from the local part unless
4470 rcpt_include_affixes is set on the transport.
4472 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4473 timeout during the DATA phase of an incoming message.
4475 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4476 applied a patch sanctioned by the Sieve maintainer.
4478 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4479 the uid or gid is negative. A case of a negative gid caused this to be
4480 noticed. The fix allows for either to be negative.
4482 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4483 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4486 PH/36 The change PH/12 above was broken. Fixed it.
4488 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4489 the grounds that routing the same address twice would always produce the
4490 same answer. This might have been true once, but it is certainly no
4491 longer true now. Routing a child address may depend on the previous
4492 routing that produced that child. Some complicated redirection strategies
4493 went wrong when messages had multiple recipients, and made Exim's
4494 behaviour dependent on the order in which the addresses were given.
4496 I have moved the duplicate checking until after the routing is complete.
4497 Exim scans the addresses that are assigned to local and remote
4498 transports, and removes any duplicates. This means that more work will be
4499 done, as duplicates will always all be routed, but duplicates are
4500 presumably rare, so I don't expect this is of any significance.
4502 For deliveries to pipes, files, and autoreplies, the duplicate checking
4503 still happens during the routing process, since they are not going to be
4506 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4507 It corrects a timeout issue with spamd. This is Ian's comment: "The
4508 background is that sometimes spamd either never reads data from a
4509 connection it has accepted, or it never writes response data. The exiscan
4510 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4511 blindly assumes that writes won't block so it may never time out."
4513 PH/39 Allow G after quota size as well as K and M.
4515 PH/40 The value set for $authenticated_id in an authenticator may not contain
4516 binary zeroes or newlines because the value is written to log lines and
4517 to spool files. There was no check on this. Now the value is run through
4518 the string_printing() function so that such characters are converted to
4519 printable escape sequences.
4521 PH/41 $message_linecount is a new variable that contains the total number of
4522 lines in the message. Compare $body_linecount, which is the count for the
4525 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4526 bounce and delay warning messages, except in certain special cases, which
4529 (a) An SMTP error message from a remote host;
4530 (b) A message specified in a :fail: redirection;
4531 (c) A message specified in a "fail" command in a system filter;
4532 (d) A message specified in a FAIL return from the queryprogram router;
4533 (e) A message specified by the cannot_route_message router option.
4535 In these cases only, Exim does include the error details in bounce and
4536 warning messages. There are also a few cases where bland messages such
4537 as "unrouteable address" or "local delivery error" are given.
4539 PH/43 $value is now also set for the "else" part of a ${run expansion.
4541 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4542 being worked on, but at least Exim now implements the latest version to
4545 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4546 process to complete was treated as a delivery failure, a timeout while
4547 writing the message to the pipe was logged, but erroneously treated as a
4548 successful delivery. Such timeouts include transport filter timeouts. For
4549 consistency with the overall process timeout, these timeouts are now
4550 treated as errors, giving rise to delivery failures by default. However,
4551 there is now a new Boolean option for the pipe transport called
4552 timeout_defer, which, if set TRUE, converts the failures into defers for
4553 both kinds of timeout. A transport filter timeout is now identified in
4556 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4557 systems where "make" and "gmake" are different, calling "gmake" at top
4558 level broke things. I've arranged for the value of $(MAKE) to be passed
4559 from the Makefile to this script so that it can call the same version of
4563 A note about Exim versions 4.44 and 4.50
4564 ----------------------------------------
4566 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4567 changes of various kinds. As a consequence, a big documentation update was
4568 needed. This delayed the release for rather longer than seemed good, especially
4569 in the light of a couple of (minor) security issues. Therefore, the changes
4570 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4571 release. So 4.44 and 4.50 are in effect two different branches that both start
4574 I have left the 4.50 change log unchanged; it contains all the changes since
4575 4.43. The change log for 4.44 is below; many of its items are identical to
4576 those for 4.50. This seems to be the most sensible way to preserve the
4577 historical information.
4583 1. Minor wording change to the doc/README.SIEVE file.
4585 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4586 computation of the current number of files was incorrect.
4588 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4589 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4592 4. Give more explanation in the error message when the command for a transport
4593 filter fails to execute.
4595 5. There are several places where Exim runs a non-Exim command in a
4596 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4597 was being done only for the command run by the queryprogram router. It is
4598 now done for all such subprocesses. The other cases are: ${run, transport
4599 filters, and the commands run by the lmtp and pipe transports.
4601 6. Added CONFIGURE_GROUP build-time option.
4603 7. Some older OS have a limit of 256 on the maximum number of file
4604 descriptors. Exim was using setrlimit() to set 1000 as a large value
4605 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4606 systems. I've change it so that if it can't get 1000, it tries for 256.
4608 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4609 was an oversight, and furthermore, ever since the addition of extra
4610 controls (e.g. 4.43/32), the checks on when to allow different forms of
4611 "control" were broken. There should now be diagnostics for all cases when a
4612 control that does not make sense is encountered.
4614 9. Added the /retain_sender option to "control=submission".
4616 10. $recipients is now available in the predata ACL (oversight).
4618 11. Tidy the search cache before the fork to do a delivery from a message
4619 received from the command line. Otherwise the child will trigger a lookup
4620 failure and thereby defer the delivery if it tries to use (for example) a
4621 cached ldap connection that the parent has called unbind on.
4623 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4624 of $address_data from the recipient verification was clobbered by the
4625 sender verification.
4627 13. The value of address_data from a sender verification is now available in
4628 $sender_address_data in subsequent conditions in the ACL statement.
4630 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4632 15. Added a new option "connect=<time>" to callout options, to set a different
4635 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4636 was its contents. (It was OK if the option was not defined at all.)
4638 17. A "Completed" log line is now written for messages that are removed from
4639 the spool by the -Mrm option.
4641 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4642 information about exactly what failed.
4644 19. Added -dd to debug only the daemon process.
4646 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4647 handles timeouts, both on the server side and network timeouts. Renamed the
4648 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4650 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4651 It is now set to "smtps".
4653 22. $host_address is now set to the target address during the checking of
4654 ignore_target_hosts.
4656 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4657 being passed; this would have caused $sender_host_name to have been used if
4658 matching the list had actually called for a host name (not very likely,
4659 since this list is usually IP addresses). A host name is now passed as
4662 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4663 code that specifies a non-restarting handler (typically sigaction() in
4664 modern systems) in an attempt to fix a rare and obscure crash bug.
4666 25. Narrowed the window for a race in the daemon that could cause it to ignore
4667 SIGCHLD signals. This is not a major problem, because they are used only to
4668 wake it up if nothing else does.
4670 26. A malformed maildirsize file could cause Exim to calculate negative values
4671 for the mailbox size or file count. Odd effects could occur as a result.
4672 The maildirsize information is now recalculated if the size or filecount
4675 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4676 support for a long time. Removed HAVE_SYS_VFS_H.
4678 28. Installed the latest version of exipick from John Jetmore.
4680 29. In an address list, if the pattern was not a regular expression, an empty
4681 subject address (from a bounce message) matched only if the pattern was an
4682 empty string. Non-empty patterns were not even tested. This was the wrong
4683 because it is perfectly reasonable to use an empty address as part of a
4684 database query. An empty address is now tested by patterns that are
4685 lookups. However, all the other forms of pattern expect the subject to
4686 contain a local part and a domain, and therefore, for them, an empty
4687 address still always fails if the pattern is not itself empty.
4689 30. Exim went into a mad DNS loop when attempting to do a callout where the
4690 host was specified on an smtp transport, and looking it up yielded more
4691 than one IP address.
4693 31. Re-factored the code for checking spool and log partition space into a
4694 function that finds that data and another that does the check. The former
4695 is then used to implement four new variables: $spool_space, $log_space,
4696 $spool_inodes, and $log_inodes.
4698 32. The RFC2047 encoding function was originally intended for short strings
4699 such as real names; it was not keeping to the 75-character limit for
4700 encoded words that the RFC imposes. It now respects the limit, and
4701 generates multiple encoded words if necessary. To be on the safe side, I
4702 have increased the buffer size for the ${rfc2047: expansion operator from
4705 33. It is now permitted to omit both strings after an "if" condition; if the
4706 condition is true, the result is "true". As before, when the second string
4707 is omitted, a false condition yields an empty string. This makes it less
4708 cumbersome to write custom ACL and router conditions.
4710 34. Failure to deliver a bounce message always caused it to be frozen, even if
4711 there was an errors_to setting on the router. The errors_to setting is now
4714 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4715 canonical form (fully expanded) before being placed in
4716 $sender_host_address.
4718 36. The table in the code that translates DNS record types into text (T_A to
4719 "A" for instance) was missing entries for NS and CNAME. It is just possible
4720 that this could have caused confusion if both these types were looked up
4721 for the same domain, because the text type is used as part of Exim's
4722 per-process caching. But the chance of anyone hitting this buglet seems
4725 37. The dnsdb lookup has been extended in a number of ways.
4727 (1) There is a new type, "zns", which walks up the domain tree until it
4728 finds some nameserver records. It should be used with care.
4730 (2) There is a new type, "mxh", which is like "mx" except that it returns
4731 just the host names, not the priorities.
4733 (3) It is now possible to give a list of domains (or IP addresses) to be
4734 looked up. The behaviour when one of the lookups defers can be
4735 controlled by a keyword.
4737 (4) It is now possible to specify the separator character for use when
4738 multiple records are returned.
4740 38. The dnslists ACL condition has been extended: it is now possible to supply
4741 a list of IP addresses and/or domains to be looked up in a particular DNS
4744 39. Added log_selector=+queue_time_overall.
4746 40. When running the queue in the test harness, wait just a tad after forking a
4747 delivery process, to get repeatability of debugging output.
4749 41. Include certificate and key file names in error message when GnuTLS fails
4750 to set them up, because the GnuTLS error message doesn't include the name
4751 of the failing file when there is a problem reading it.
4753 42. Allow both -bf and -bF in the same test run.
4755 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4757 44. The "Exiscan patch" is now merged into the mainline Exim source.
4759 45. Sometimes the final signoff response after QUIT could fail to get
4760 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4761 < 0 before doing a fflush(). This bug looks as though it goes back to the
4762 introduction of TLS in release 3.20, but "sometimes" must have been rare
4763 because the tests only now provoked it.
4765 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4766 (this can affect the format of dates).
4768 47. exim_tidydb, when checking for the continued existence of a message for
4769 which it has found a message-specific retry record, was not finding
4770 messages that were in split spool directories. Consequently, it was
4771 deleting retry records that should have stayed in existence.
4773 48. Steve fixed some bugs in eximstats.
4775 49. The SPA authentication driver was not abandoning authentication and moving
4776 on to the next authenticator when an expansion was forced to fail,
4777 contradicting the general specification for all authenticators. Instead it
4778 was generating a temporary error. It now behaves as specified.
4780 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4781 (the order specifies the preference for clients). The order is now AES256,
4782 AES128, 3DES, ARCFOUR128.
4784 51. Small patch to Sieve code - explicitly set From: when generating an
4787 52. Exim crashed if a remote delivery caused a very long error message to be
4788 recorded - for instance if somebody sent an entire SpamAssassin report back
4789 as a large number of 550 error lines. This bug was coincidentally fixed by
4790 increasing the size of one of Exim's internal buffers (big_buffer) that
4791 happened as part of the Exiscan merge. However, to be on the safe side, I
4792 have made the code more robust (and fixed the comments that describe what
4795 53. Now that there can be additional text after "Completed" in log lines (if
4796 the queue_time_overall log selector is set), a one-byte patch to exigrep
4797 was needed to allow it to recognize "Completed" as not the last thing in
4800 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4801 patch that reportedly fixes this has been added. I am not expert enough to
4802 create a test for it. This is what the patch creator wrote:
4804 "I found a little strange behaviour of ldap code when working with
4805 Windows 2003 AD Domain, where users was placed in more than one
4806 Organization Units. When I tried to give exim partial DN, the exit code
4807 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4808 But simultaneously result of request was absolutely normal ldap result,
4809 so I produce this patch..."
4811 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4812 REFERENCE, so I have modified the code to exclude the patch when that macro
4815 55. Some experimental protocols are using DNS PTR records for new purposes. The
4816 keys for these records are domain names, not reversed IP addresses. The
4817 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4818 leaves it alone. Component reversal etc. now happens only for IP addresses.
4821 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4823 57. Double the size of the debug message buffer (to 2048) so that more of very
4824 long debug lines gets shown.
4826 58. The exicyclog utility now does better if the number of log files to keep
4827 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4829 59. Two changes related to the smtp_active_hostname option:
4831 (1) $smtp_active_hostname is now available as a variable.
4832 (2) The default for smtp_banner uses $smtp_active_hostname instead
4833 of $primary_hostname.
4835 60. The host_aton() function is supposed to be passed a string that is known
4836 to be a valid IP address. However, in the case of IPv6 addresses, it was
4837 not checking this. This is a hostage to fortune. Exim now panics and dies
4838 if the condition is not met. A case was found where this could be provoked
4839 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4840 components; fortuitously, this particular loophole had already been fixed
4841 by change 4.50/55 above.
4843 If there are any other similar loopholes, the new check in host_aton()
4844 itself should stop them being exploited. The report I received stated that
4845 data on the command line could provoke the exploit when Exim was running as
4846 exim, but did not say which command line option was involved. All I could
4847 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4848 running as the user.
4851 61. There was a buffer overflow vulnerability in the SPA authentication code
4852 (which came originally from the Samba project). I have added a test to the
4853 spa_base64_to_bits() function which I hope fixes it.
4856 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4857 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4859 63. The daemon start-up calls getloadavg() while still root for those OS that
4860 need the first call to be done as root, but it missed one case: when
4861 deliver_queue_load_max is set with deliver_drop_privilege. This is
4862 necessary for the benefit of the queue runner, because there is no re-exec
4863 when deliver_drop_privilege is set.
4865 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4866 This has been fixed.
4868 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4869 was in use, was not putting the data itself into the right store pool;
4870 consequently, it could be overwritten for a subsequent message in the same
4871 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4874 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4876 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4877 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4878 the function were treating the return as a boolean value, which happened to
4879 work because 0=false and not-0=true, but is not correct code.
4881 68. The host_aton() function was not handling scoped IPv6 addresses (those
4882 with, for example, "%eth0" on the end) correctly.
4884 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4885 negated items (that is, ~something) in unsigned ints. Some compilers
4886 apparently mutter when there is no cast.
4888 70. If an address verification called from an ACL failed, and did not produce a
4889 user-specific message (i.e. there was only a "system" message), nothing was
4890 put in $acl_verify_message. In this situation, it now puts the system
4893 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4894 session; change 4.31/43 added the unwanted input to the log line - except
4895 that it did not do this in the start of session case. It now does.
4897 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4898 This is wrong and can cause the other end to generate a synchronization
4899 error if it is another Exim or anything else that does the synchronization
4900 check. A QUIT command is no longer sent after a timeout.
4902 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4903 during host lookups.
4905 74. The defer_ok option of callout verification was not working if it was used
4906 when verifying addresses in header lines, that is, for this case:
4908 verify = header_sender/callout=defer_ok
4910 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4911 those file descriptors could be used for SMTP connections. If anything
4912 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4913 could be sent to the SMTP client, causing chaos. The daemon now opens
4914 stdin, stdout, and stderr to /dev/null when it puts itself into the
4917 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4918 log by default. The user can override this with suitable Perl magic.
4920 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4921 the log message when discard triggers, was not working for the DATA ACL or
4922 for the non-SMTP ACL.
4924 78. Error message wording change in sieve.c.
4926 79. If smtp_accept_max_per_host was set, the number of connections could be
4927 restricted to fewer than expected, because the daemon was trying to set up
4928 a new connection before checking whether the processes handling previous
4929 connections had finished. The check for completed processes is now done
4930 earlier. On busy systems, this bug wouldn't be noticed because something
4931 else would have woken the daemon, and it would have reaped the completed
4934 80. If a message was submitted locally by a user whose login name contained one
4935 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4936 It caused a spool format error. I have fixed the spool reading code. A
4937 related problem was that the "from" clause in the Received: line became
4938 illegal because of the space(s). It is now covered by ${quote_local_part.
4940 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4943 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4945 83. Patch from Sieve maintainer to fix unterminated string problem in
4946 "vacation" handling.
4948 84. Some minor changes to the Linux configuration files to help with other
4949 OS variants using glibc.
4951 85. One more patch for Sieve to update vacation handling to latest spec.
4954 ----------------------------------------------------
4955 See the note above about the 4.44 and 4.50 releases.
4956 ----------------------------------------------------
4962 1. Change 4.43/35 introduced a bug that caused file counts to be
4963 incorrectly computed when quota_filecount was set in an appendfile
4966 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4967 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4970 3. Give more explanation in the error message when the command for a transport
4971 filter fails to execute.
4973 4. There are several places where Exim runs a non-Exim command in a
4974 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4975 was being done only for the command run by the queryprogram router. It is
4976 now done for all such subprocesses. The other cases are: ${run, transport
4977 filters, and the commands run by the lmtp and pipe transports.
4979 5. Some older OS have a limit of 256 on the maximum number of file
4980 descriptors. Exim was using setrlimit() to set 1000 as a large value
4981 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4982 systems. I've change it so that if it can't get 1000, it tries for 256.
4984 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4985 was an oversight, and furthermore, ever since the addition of extra
4986 controls (e.g. 4.43/32), the checks on when to allow different forms of
4987 "control" were broken. There should now be diagnostics for all cases when a
4988 control that does not make sense is encountered.
4990 7. $recipients is now available in the predata ACL (oversight).
4992 8. Tidy the search cache before the fork to do a delivery from a message
4993 received from the command line. Otherwise the child will trigger a lookup
4994 failure and thereby defer the delivery if it tries to use (for example) a
4995 cached ldap connection that the parent has called unbind on.
4997 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4998 of $address_data from the recipient verification was clobbered by the
4999 sender verification.
5001 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5002 was its contents. (It was OK if the option was not defined at all.)
5004 11. A "Completed" log line is now written for messages that are removed from
5005 the spool by the -Mrm option.
5007 12. $host_address is now set to the target address during the checking of
5008 ignore_target_hosts.
5010 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5011 being passed; this would have caused $sender_host_name to have been used if
5012 matching the list had actually called for a host name (not very likely,
5013 since this list is usually IP addresses). A host name is now passed as
5016 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5017 code that specifies a non-restarting handler (typically sigaction() in
5018 modern systems) in an attempt to fix a rare and obscure crash bug.
5020 15. Narrowed the window for a race in the daemon that could cause it to ignore
5021 SIGCHLD signals. This is not a major problem, because they are used only to
5022 wake it up if nothing else does.
5024 16. A malformed maildirsize file could cause Exim to calculate negative values
5025 for the mailbox size or file count. Odd effects could occur as a result.
5026 The maildirsize information is now recalculated if the size or filecount
5029 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5030 support for a long time. Removed HAVE_SYS_VFS_H.
5032 18. Updated exipick to current release from John Jetmore.
5034 19. Allow an empty sender to be matched against a lookup in an address list.
5035 Previously the only cases considered were a regular expression, or an
5038 20. Exim went into a mad DNS lookup loop when doing a callout where the
5039 host was specified on the transport, if the DNS lookup yielded more than
5042 21. The RFC2047 encoding function was originally intended for short strings
5043 such as real names; it was not keeping to the 75-character limit for
5044 encoded words that the RFC imposes. It now respects the limit, and
5045 generates multiple encoded words if necessary. To be on the safe side, I
5046 have increased the buffer size for the ${rfc2047: expansion operator from
5049 22. Failure to deliver a bounce message always caused it to be frozen, even if
5050 there was an errors_to setting on the router. The errors_to setting is now
5053 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5054 canonical form (fully expanded) before being placed in
5055 $sender_host_address.
5057 24. Updated eximstats to version 1.33
5059 25. Include certificate and key file names in error message when GnuTLS fails
5060 to set them up, because the GnuTLS error message doesn't include the name
5061 of the failing file when there is a problem reading it.
5063 26. Expand error message when OpenSSL has problems setting up cert/key files.
5066 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5067 (this can affect the format of dates).
5069 28. exim_tidydb, when checking for the continued existence of a message for
5070 which it has found a message-specific retry record, was not finding
5071 messages that were in split spool directories. Consequently, it was
5072 deleting retry records that should have stayed in existence.
5074 29. eximstats updated to version 1.35
5075 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5076 1.35 - bugfix such that pie charts by volume are generated correctly
5078 30. The SPA authentication driver was not abandoning authentication and moving
5079 on to the next authenticator when an expansion was forced to fail,
5080 contradicting the general specification for all authenticators. Instead it
5081 was generating a temporary error. It now behaves as specified.
5083 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5084 (the order specifies the preference for clients). The order is now AES256,
5085 AES128, 3DES, ARCFOUR128.
5087 31. Small patch to Sieve code - explicitly set From: when generating an
5090 32. Exim crashed if a remote delivery caused a very long error message to be
5091 recorded - for instance if somebody sent an entire SpamAssassin report back
5092 as a large number of 550 error lines. This bug was coincidentally fixed by
5093 increasing the size of one of Exim's internal buffers (big_buffer) that
5094 happened as part of the Exiscan merge. However, to be on the safe side, I
5095 have made the code more robust (and fixed the comments that describe what
5098 33. Some experimental protocols are using DNS PTR records for new purposes. The
5099 keys for these records are domain names, not reversed IP addresses. The
5100 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5101 leaves it alone. Component reversal etc. now happens only for IP addresses.
5104 34. The host_aton() function is supposed to be passed a string that is known
5105 to be a valid IP address. However, in the case of IPv6 addresses, it was
5106 not checking this. This is a hostage to fortune. Exim now panics and dies
5107 if the condition is not met. A case was found where this could be provoked
5108 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5109 components; fortuitously, this particular loophole had already been fixed
5110 by change 4.50/55 or 4.44/33 above.
5112 If there are any other similar loopholes, the new check in host_aton()
5113 itself should stop them being exploited. The report I received stated that
5114 data on the command line could provoke the exploit when Exim was running as
5115 exim, but did not say which command line option was involved. All I could
5116 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5117 running as the user.
5120 35. There was a buffer overflow vulnerability in the SPA authentication code
5121 (which came originally from the Samba project). I have added a test to the
5122 spa_base64_to_bits() function which I hope fixes it.
5125 36. The daemon start-up calls getloadavg() while still root for those OS that
5126 need the first call to be done as root, but it missed one case: when
5127 deliver_queue_load_max is set with deliver_drop_privilege. This is
5128 necessary for the benefit of the queue runner, because there is no re-exec
5129 when deliver_drop_privilege is set.
5131 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5132 was in use, was not putting the data itself into the right store pool;
5133 consequently, it could be overwritten for a subsequent message in the same
5134 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5137 38. Sometimes the final signoff response after QUIT could fail to get
5138 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5139 < 0 before doing a fflush(). This bug looks as though it goes back to the
5140 introduction of TLS in release 3.20, but "sometimes" must have been rare
5141 because the tests only now provoked it.
5147 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5148 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5149 Now it returns an integer. A number of calls were still expecting a T/F
5150 return. Fortuitously, in all cases, the tests worked in OK situations,
5151 which is the norm. However, things would have gone wrong on any write
5152 failures on the smtp file descriptor. This function is used when sending
5153 messages over SMTP and also when doing verify callouts.
5155 2. When Exim is called to do synchronous delivery of a locally submitted
5156 message (the -odf or -odi options), it no longer closes stderr before doing
5159 3. Implemented the mua_wrapper option.
5161 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5163 5. Implemented the functions header_remove(), header_testname(),
5164 header_add_at_position(), and receive_remove_recipient(), and exported them
5167 6. If an ACL "warn" statement specified the addition of headers, Exim already
5168 inserted X-ACL-Warn: at the start if there was no header name. However, it
5169 was not making this test for the second and subsequent header lines if
5170 there were newlines in the string. This meant that an invalid header could
5171 be inserted if Exim was badly configured.
5173 7. Allow an ACL "warn" statement to add header lines at the start or after all
5174 the Received: headers, as well as at the end.
5176 8. Added the rcpt_4xx retry error code.
5178 9. Added postmaster_mailfrom=xxx to callout verification option.
5180 10. Added mailfrom=xxxx to the callout verification option, for verify=
5183 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5184 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5186 12. Inserted some casts to stop certain compilers warning when using pointer
5187 differences as field lengths or precisions in printf-type calls (mostly
5188 affecting debugging statements).
5190 13. Added optional readline() support for -be (dynamically loaded).
5192 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5193 same clock tick as a message's arrival, so that its received time was the
5194 same as the "first fail" time on the retry record, and that message
5195 remained on the queue past the ultimate address timeout, every queue runner
5196 would try a delivery (because it was past the ultimate address timeout) but
5197 after another failure, the ultimate address timeout, which should have then
5198 bounced the address, did not kick in. This was a "< instead of <=" error;
5199 in most cases the first failure would have been in the next clock tick
5200 after the received time, and all would be well.
5202 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5203 being recognized when the domain list was tested by the match_domain
5204 condition in an expansion string.
5206 16. Added the ${str2b64: operator.
5208 17. Exim was always calling setrlimit() to set a large limit for the number of
5209 processes, without checking whether the existing limit was already
5210 adequate. (It did check for the limit on file descriptors.) Furthermore,
5211 errors from getrlimit() and setrlimit() were being ignored. Now they are
5212 logged to the main and panic logs, but Exim does carry on, to try to do its
5213 job under whatever limits there are.
5215 18. Imported PCRE 5.0.
5217 19. Trivial typo in log message " temporarily refused connection" (the leading
5220 20. If the log selector return_path_on_delivery was set and an address was
5221 redirected to /dev/null, the delivery process crashed because it assumed
5222 that a return path would always be set for a "successful" delivery. In this
5223 case, the whole delivery is bypassed as an optimization, and therefore no
5226 21. Internal re-arrangement: the function for sending a challenge and reading
5227 a response while authentication was assuming a zero-terminated challenge
5228 string. It's now changed to take a pointer and a length, to allow for
5229 binary data in such strings.
5231 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5233 23. Exim was not respecting finduser_retries when seeking the login of the
5234 uid under which it was called; it was always trying 10 times. (The default
5235 setting of finduser_retries is zero.) Also, it was sleeping after the final
5236 failure, which is pointless.
5238 24. Implemented tls_on_connect_ports.
5240 25. Implemented acl_smtp_predata.
5242 26. If the domain in control=submission is set empty, Exim assumes that the
5243 authenticated id is a complete email address when it generates From: or
5244 Sender: header lines.
5246 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5247 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5248 chown and chgrp in /bin and hostname in /usr/bin.
5250 28. Exim was keeping the "process log" file open after each use, just as it
5251 does for the main log. This opens the possibility of it remaining open for
5252 long periods when the USR1 signal hits a daemon. Occasional processlog
5253 errors were reported, that could have been caused by this. Anyway, it seems
5254 much more sensible not to leave this file open at all, so that is what now
5257 29. The long-running daemon process does not normally write to the log once it
5258 has entered its main loop, and it closes the log before doing so. This is
5259 so that log files can straightforwardly be renamed and moved. However,
5260 there are a couple of unusual error situations where the daemon does write
5261 log entries, and I had neglected to close the log afterwards.
5263 30. The text of an SMTP error response that was received during a remote
5264 delivery was being truncated at 512 bytes. This is too short for some of
5265 the long messages that one sometimes sees. I've increased the limit to
5268 31. It is now possible to make retry rules that apply only when a message has a
5269 specific sender, in particular, an empty sender.
5271 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5272 it possible to be selective about when SMTP synchronization is enforced.
5274 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5276 32. Implemented hosts_connection_nolog.
5278 33. Added an ACL for QUIT.
5280 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5283 35. Added mailbox_size and mailbox_filecount to appendfile.
5285 36. Added control = no_multiline_responses to ACLs.
5287 37. There was a bug in the logic of the code that waits for the clock to tick
5288 in the case where the clock went backwards by a substantial amount such
5289 that the microsecond fraction of "now" was more than the microsecond
5290 fraction of "then" (but the whole seconds number was less).
5292 38. Added support for the libradius Radius client library this is found on
5293 FreeBSD (previously only the radiusclient library was supported).
5299 1. When certain lookups returned multiple values in the form name=value, the
5300 quoting of the values was not always being done properly. Specifically:
5301 (a) If the value started with a double quote, but contained no whitespace,
5303 (b) If the value contained whitespace other than a space character (i.e.
5304 tabs or newlines or carriage returns) it was not quoted.
5305 This fix has been applied to the mysql and pgsql lookups by writing a
5306 separate quoting function and calling it from the lookup code. The fix
5307 should probably also be applied to nisplus, ibase and oracle lookups, but
5308 since I cannot test any of those, I have not disturbed their existing code.
5310 2. A hit in the callout cache for a specific address caused a log line with no
5311 reason for rejecting RCPT. Now it says "Previous (cached) callout
5312 verification failure".
5314 3. There was an off-by-one bug in the queryprogram router. An over-long
5315 return line was truncated at 256 instead of 255 characters, thereby
5316 overflowing its buffer with the terminating zero. As well as fixing this, I
5317 have increased the buffer size to 1024 (and made a note to document this).
5319 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5320 when Exim is waiting for an SMTP response from a remote server, Exim
5321 restarts its select() call on the socket, thereby resetting its timeout.
5322 This is not a problem when such interrupts are rare. Somebody set up a cron
5323 job to run exiwhat every 2 minutes, which is less than the normal select()
5324 timeout (5 or 10 minutes). This meant that the select() timeout never
5325 kicked in because it was always reset. I have fixed this by comparing the
5326 time when an interrupt arrives with the time at the start of the first call
5327 to select(). If more time than the timeout has elapsed, the interrupt is
5328 treated as a timeout.
5330 5. Some internal re-factoring in preparation for the addition of Sieve
5331 extensions (by MH). In particular, the "personal" test is moved to a
5332 separate function, and given an option for scanning Cc: and Bcc: (which is
5333 not set for Exim filters).
5335 6. When Exim created an email address using the login of the caller as the
5336 local part (e.g. when creating a From: or Sender: header line), it was not
5337 quoting the local part when it contained special characters such as @.
5339 7. Installed new OpenBSD configuration files.
5341 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5342 try to make them clearer.
5344 9. Callout options, other than the timeout value, were being ignored when
5345 verifying sender addresses in header lines. For example, when using
5347 verify = header_sender/callout=no_cache
5349 the cache was (incorrectly) being used.
5351 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5352 only the Cygwin environment.
5354 11. When return_path_on_delivery was set as a log selector, if different remote
5355 addresses in the same message used different return paths and parallel
5356 remote delivery occurred, the wrong values would sometimes be logged.
5357 (Whenever a remote delivery process finished, the return path value from
5358 the most recently started remote delivery process was logged.)
5360 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5361 lines when AUTH and/or TLS are in use. This is the "received protocol"
5362 field. Exim used to use "asmtp" for authenticated SMTP, without any
5363 indication (in the protocol name) for TLS use. Now it follows the RFC and
5364 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5365 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5366 names appear in log lines as well as in Received: header lines.
5368 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5369 extensions, and comparison tests, and to fix some bugs.
5371 14. Changes to the "personal" filter test:
5373 (1) The test was buggy in that it was just doing the equivalent of
5374 "contains" tests on header lines. For example, if a user's address was
5375 anne@some.where, the "personal" test would incorrectly be true for
5377 To: susanne@some.where
5379 This test is now done by extracting each address from the header in turn,
5380 and checking the entire address. Other tests that are part of "personal"
5381 are now done using regular expressions (for example, to check local parts
5382 of addresses in From: header lines).
5384 (2) The list of non-personal local parts in From: addresses has been
5385 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5386 taken from the Sieve specification recommendations.
5388 (3) If the message contains any header line starting with "List-" it is
5389 treated as non-personal.
5391 (4) The test for "circular" in the Subject: header line has been removed
5392 because it now seems ill-conceived.
5394 15. Minor typos in src/EDITME comments corrected.
5396 16. Installed latest exipick from John Jetmore.
5398 17. If headers_add on a router specified a text string that was too long for
5399 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5400 of string_sprintf() is now avoided.
5402 18. $message_body_size was not set (it was always zero) when running the DATA
5403 ACL and the local_scan() function.
5405 19. For the "mail" command in an Exim filter, no default was being set for
5406 the once_repeat time, causing a random time value to be used if "once" was
5407 specified. (If the value happened to be <= 0, no repeat happened.) The
5408 default is now 0s, meaning "never repeat". The "vacation" command was OK
5409 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5410 (I found it when inspecting the code).
5412 20. There is now an overall timeout for performing a callout verification. It
5413 defaults to 4 times the callout timeout, which applies to individual SMTP
5414 commands during the callout. The overall timeout applies when there is more
5415 than one host that can be tried. The timeout is checked before trying the
5416 next host. This prevents very long delays if there are a large number of
5417 hosts and all are timing out (e.g. when the network connections are timing
5418 out). The value of the overall timeout can be changed by specifying an
5419 additional sub-option for "callout", called "maxwait". For example:
5421 verify = sender/callout=5s,maxwait=20s
5423 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5424 to the end before writing, but this should make it even safer).
5426 22. Exim was forgetting that it had advertised PIPELINING for the second and
5427 subsequent messages on an SMTP connection. It was also not resetting its
5428 memory on STARTTLS and an internal HELO.
5430 23. When Exim logs an SMTP synchronization error within a session, it now
5431 records whether PIPELINING has been advertised or not.
5433 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5434 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5435 rather than long int.
5437 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5439 26. Added the never_mail option to autoreply.
5445 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5446 crash if the getsockname() call failed; this can happen if a connection is
5447 closed very soon after it is established. The problem was simply in the
5448 order in which certain operations were done, causing Exim to try to write
5449 to the SMTP stream before it had set up the file descriptor. The bug has
5450 been fixed by making things happen in the correct order.
5456 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5457 before the connection was closed, thus losing the rejection response.
5459 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5460 some early Solaris releases, but causes trouble in current releases where
5461 socklen_t is defined.
5463 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5466 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5469 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5470 that releasing the top of it at the end releases what was used for sub-
5471 expansions (unless the block got too big). However, discard this block if
5472 the first thing is a variable or header, so that we can use its block when
5473 it is dynamic (useful for very large $message_headers, for example).
5475 6. Lookups now cache *every* query, not just the most recent. A new, separate
5476 store pool is used for this. It can be recovered when all lookup caches are
5477 flushed. Lookups now release memory at the end of their result strings.
5478 This has involved some general refactoring of the lookup sources.
5480 7. Some code has been added to the store_xxx() functions to reduce the amount
5481 of flapping under certain conditions.
5483 8. log_incoming_interface used to affect only the <= reception log lines. Now
5484 it causes the local interface and port to be added to several more SMTP log
5485 lines, for example "SMTP connection from", and rejection lines.
5487 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5489 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5491 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5492 could be overwritten at the end of the current message (or the start of a
5493 new message if it was set in a HELO ACL). The value is now preserved for
5494 the duration of the SMTP connection.
5496 12. If a transport had a headers_rewrite setting, and a matching header line
5497 contained an unqualified address, that address was qualified, even if it
5498 did not match any rewriting rules. The underlying bug was that the values
5499 of the flags that permit the existence of unqualified sender and recipient
5500 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5501 non-local messages, and by -bnq for local messages) were not being
5502 preserved with the message after it was received.
5504 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5505 "next input=" as part of the text comprising the host identity instead of
5506 the correct text. The code was using the same buffer for two different
5507 strings. However, depending on which order the printing function evaluated
5508 its arguments, the bug did not always show up. Under Linux, for example, my
5509 test suite worked just fine.
5511 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5512 doesn't work with some older versions of Perl. It has been changed to "my",
5513 which in any case is probably the better facility to use.
5515 15. A really picky compiler found some instances of statements for creating
5516 error messages that either had too many or two few arguments for the format
5519 16. The size of the buffer for calls to the DNS resolver has been increased
5520 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5521 for addresses that have a lot of PTR records. This alleviates a problem; it
5522 does not fully solve it.
5524 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5525 buffer now truncates the list and logs the incident, which is the same
5526 action as happens when Exim is looking up a host name and its aliases.
5527 Previously in this situation something unpredictable would happen;
5528 sometimes it was "internal error: store_reset failed".
5530 18. If a server dropped the connection unexpectedly when an Exim client was
5531 using GnuTLS and trying to read a response, the client delivery process
5532 crashed while trying to generate an error log message.
5534 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5535 string, for example:
5537 warn message = H1: something\nH2: something
5539 the text was added as a single header line from Exim's point of view
5540 though it ended up OK in the delivered message. However, searching for the
5541 second and subsequent header lines using $h_h2: did not work. This has been
5542 fixed. Similarly, if a system filter added multiple headers in this way,
5543 the routers could not see them.
5545 20. Expanded the error message when iplsearch is called with an invalid key to
5546 suggest using net-iplsearch in a host list.
5548 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5549 ACLs are no longer actually imposed (and a message to that effect is
5552 22. If a "gecos" field in a passwd entry contained escaped characters, in
5553 particular, if it contained a \" sequence, Exim got it wrong when building
5554 a From: or a Sender: header from that name. A second bug also caused
5555 incorrect handling when an unquoted " was present following a character
5556 that needed quoting.
5558 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5559 was not being matched caselessly.
5561 24. Arranged for all hyphens in the exim.8 source to be escaped with
5564 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5565 properly. Recipient callout cache records were still being keyed to include
5566 the sender, even when use_sender was set false. This led to far more
5567 callouts that were necessary. The sender is no longer included in the key
5568 when use_sender is false.
5570 26. Added "control = submission" modifier to ACLs.
5572 27. Added the ${base62d: operator to decode base 62 numbers.
5574 28. dnsdb lookups can now access SRV records.
5576 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5577 the configuration file.
5579 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5580 (-v) mode. This makes the output for a verbose queue run more intelligible.
5582 31. Added a use_postmaster feature to recipient callouts.
5584 32. Added the $body_zerocount variable, containing the number of binary zero
5585 bytes in the message body.
5587 33. The time of last modification of the "new" subdirectory is now used as the
5588 "mailbox time last read" when there is a quota error for a maildir
5591 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5593 35. Added +ignore_unknown as a special item in host lists.
5595 36. Code for decoding IPv6 addresses in host lists is now included, even if
5596 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5597 address was recognized as an IP address, but was then not correctly decoded
5598 into binary, causing unexpected and incorrect effects when compared with
5605 1. Very minor rewording of debugging text in manualroute to say "list of
5606 hosts" instead of "hostlist".
5608 2. If verify=header_syntax was set, and a header line with an unqualified
5609 address (no domain) and a large number of spaces between the end of the
5610 name and the colon was received, the reception process suffered a buffer
5611 overflow, and (when I tested it) crashed. This was caused by some obsolete
5612 code that should have been removed. The fix is to remove it!
5614 3. When running in the test harness, delay a bit after writing a bounce
5615 message to get a bit more predictability in the log output.
5617 4. Added a call to search_tidyup() just before forking a reception process. In
5618 theory, someone could use a lookup in the expansion of smtp_accept_max_
5619 per_host which, without the tidyup, could leave open a database connection.
5621 5. Added the variables $recipient_data and $sender_data which get set from a
5622 lookup success in an ACL "recipients" or "senders" condition, or a router
5623 "senders" option, similar to $domain_data and $local_part_data.
5625 6. Moved the writing of debug_print from before to after the "senders" test
5628 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5629 problems for message scanning, either using a data ACL, or using
5630 local_scan() because the Received: header was not generated till after they
5631 were called (in order to set the time as the time of reception completion).
5632 I have revised the way this works. The header is now generated after the
5633 body is received, but before the ACL or local_scan() are called. After they
5634 are run, the timestamp in the header is updated.
5640 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5641 before starting a queue runner without re-exec. This happened only when
5642 deliver_drop_privilege was set or when the Exim user was set to root. The
5643 effect of the bug was that timeouts during subsequent deliveries caused
5644 crashes instead of being properly handled. The handler is now left at its
5645 default (and expected) setting.
5647 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5648 message, again when deliver_drop_privilege is set or Exim is run as root.
5649 The bug described in (1) was not present in this case, but the tidying up
5650 of the other signals was missing. I have made the two cases consistent.
5652 3. The ignore_target_hosts setting on a manualroute router was being ignored
5653 for hosts that were looked up using the /MX notation.
5655 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5658 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5659 operated on the sender address. After changing the $sender_address to <>
5660 for the sender address verify, Exim was re-instated it as the original
5661 (before rewriting) address, but remembering that it had rewritten it, so it
5662 wasn't rewriting it again. This bug also had the effect of breaking the
5663 sender address verification caching when the sender address was rewritten.
5665 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5666 This has been changed so that if the ip literal address matches
5667 ignore_target_hosts, the router declines.
5669 7. Added expansion conditions match_domain, match_address, and match_local_
5670 part (NOT match_host).
5672 8. The placeholder for the Received: header didn't have a length field set.
5674 9. Added code to Exim itself and to exim_lock to test for a specific race
5675 condition that could lead to file corruption when using MBX delivery. The
5676 issue is with the lockfile that is created in /tmp. If this file is removed
5677 after a process has opened it but before that process has acquired a lock,
5678 there is the potential for a second process to recreate the file and also
5679 acquire a lock. This could lead to two Exim processes writing to the file
5680 at the same time. The added code performs the same test as UW imapd; it
5681 checks after acquiring the lock that its file descriptor still refers to
5682 the same named file.
5684 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5685 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5688 11. Added the smtp_active_hostname option. If used, this will typically be made
5689 to depend on the incoming interface address. Because $interface_address is
5690 not set up until the daemon has forked a reception process, error responses
5691 that can happen earlier (such as "too many connections") no longer contain
5694 12. If an expansion in a condition on a "warn" statement fails because a lookup
5695 defers, the "warn" statement is abandoned, and the next ACL statement is
5696 processed. Previously this caused the whole ACL to be aborted.
5698 13. Added the iplsearch lookup type.
5700 14. Added ident_timeout as a log selector.
5702 15. Added tls_certificate_verified as a log selector.
5704 16. Added a global option tls_require_ciphers (compare the smtp transport
5705 option of the same name). This controls incoming TLS connections.
5707 17. I finally figured out how to make tls_require_ciphers do a similar thing
5708 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5709 before starting the TLS session.
5711 18. Tabs are now shown as \t in -bP output.
5713 19. If the log selector return_path_on_delivery was set, Exim crashed when
5714 bouncing a message because it had too many Received: header lines.
5716 20. If two routers both had headers_remove settings, and the first one included
5717 a superfluous trailing colon, the final name in the first list and the
5718 first name in the second list were incorrectly joined into one item (with a
5719 colon in the middle).
5725 1. Added -C and -D options to the exinext utility, mainly to make it easier
5726 to include in the automated testing, but these could be helpful when
5727 multiple configurations are in use.
5729 2. The exinext utility was not formatting the output nicely when there was
5730 an alternate port involved in the retry record key, nor when there was a
5731 message id as well (for retries that were specific to a specific message
5732 and a specific host). It was also confused by IPv6 addresses, because of
5733 the additional colons they contain. I have fixed the IPv4 problem, and
5734 patched it up to do a reasonable job for IPv6.
5736 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5737 delivery, the log line now contains "pipelined" if PIPELINING was used.
5739 4. An SMTP transport process used to panic and die if the bind() call to set
5740 an explicit outgoing interface failed. This has been changed; it is now
5741 treated in the same way as a connect() failure.
5743 5. A reference to $sender_host_name in the part of a conditional expansion
5744 that was being skipped was still causing a DNS lookup. This no longer
5747 6. The def: expansion condition was not recognizing references to header lines
5748 that used bh_ and bheader_.
5750 7. Added the _cache feature to named lists.
5752 8. The code for checking quota_filecount in the appendfile transport was
5753 allowing one more file than it should have been.
5755 9. For compatibility with Sendmail, the command line option
5763 and sets the incoming protocol and host name (for trusted callers). The
5764 host name and its colon can be omitted when only the protocol is to be set.
5765 Note the Exim already has two private options, -pd and -ps, that refer to
5766 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5767 "s", but I don't think that's a major issue.
5769 10. A number of refactoring changes to the code, none of which should affect
5772 (a) The number of logging options was getting close to filling up the
5773 32-bit word that was used as a bit map. I have split them into two classes:
5774 those that are passed in the argument to log_write(), and those that are
5775 only ever tested independently outside of that function. These are now in
5776 separate 32-bit words, so there is plenty of room for expansion again.
5777 There is no change in the user interface or the logging behaviour.
5779 (b) When building, for example, log lines, the code previously used a
5780 macro that called string_cat() twice, in order to add two strings. This is
5781 not really sufficiently general. Furthermore, there was one instance where
5782 it was actually wrong because one of the argument was used twice, and in
5783 one call a function was used. (As it happened, calling the function twice
5784 did not affect the overall behaviour.) The macro has been replaced by a
5785 function that can join an arbitrary number of extra strings onto a growing
5788 (c) The code for expansion conditions now uses a table and a binary chop
5789 instead of a serial search (which was left over from when there were very
5790 few conditions). Also, it now recognizes conditions like "pam" even when
5791 the relevant support is not compiled in: a suitably worded error message is
5792 given if an attempt is made to use such a condition.
5794 11. Added ${time_interval:xxxxx}.
5796 12. A bug was causing one of the ddress fields not to be passed back correctly
5797 from remote delivery subprocesses. The field in question was not being
5798 subsequently used, so this caused to problems in practice.
5800 13. Added new log selectors queue_time and deliver_time.
5802 14. Might have fixed a bug in maildirsizefile handling that threw up
5803 "unexpected character" debug warnings, and recalculated the data
5804 unnecessarily. In any case, I expanded the warning message to give more
5807 15. Added the message "Restricted characters in address" to the statements in
5808 the default ACL that block characters like @ and % in local parts.
5810 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5811 Three changes have been made:
5813 (a) There was a serious bug; a negative response to MAIL caused the whole
5814 recipient domain to be cached as invalid, thereby blocking all messages
5815 to all local parts at the same domain, from all senders. This bug has
5816 been fixed. The domain is no longer cached after a negative response to
5817 MAIL if the sender used is not empty.
5819 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5822 (c) A new callout option, "use_sender" has been added for people who want
5823 the modified behaviour.
5829 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5832 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5833 indeed breaks things for older releases.
5835 3. Added additional logging to the case where there is a problem reading data
5836 from a filter that is running in a subprocess using a pipe, in order to
5837 try to track down a specific problem.
5839 4. Testing facility fudge: when running in the test harness and attempting
5840 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5841 getting "No route to host". Convert this to a timeout.
5843 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5846 6. Some OS don't have socklen_t but use size_t instead. This affects the
5847 fifth argument of getsockopt() amongst other things. This is now
5848 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5849 can be set for individual OS. I have set it for SunOS5, OSF1, and
5850 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5851 some earlier ones do not.
5853 7. Change 4.30/15 was not doing the test caselessly.
5855 8. The standard form for an IPv6 address literal was being rejected by address
5856 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5857 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5858 this, as well as the form without the "IPv6" on the front (but only when
5859 address literals are enabled, of course).
5861 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5863 10. Exim crashed if a message with an empty sender address specified by -f
5864 encountered a router with an errors_to setting. This could be provoked only
5865 by a command such as
5869 where an empty string was supplied; "<>" did not hit this bug.
5871 11. Installed PCRE release 4.5.
5873 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5874 remained set. It is now erased.
5876 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5877 times from message ids (which are base 36 rather than the normal 62).
5879 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5880 were being counted as actual protocol errors, and logged if the log
5881 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5882 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5883 and DATA following a set of rejected RCPTs do not count as protocol errors.
5884 In other words, Exim assumes they were pipelined, though this may not
5885 actually be the case. Of course, in all cases the client gets an
5886 appropriate error code.
5888 15. If a lookup fails in an ACL condition, a message about the failure may
5889 be available; it is used if testing the ACL cannot continue, because most
5890 such messages specify what the cause of the deferral is. However, some
5891 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5892 that caused an old message to be retained and used if a later statement
5893 caused a defer, replacing the real cause of the deferral.
5895 16. If an IP address had so many PTR records that the DNS lookup buffer
5896 was not large enough to hold them, Exim could crash while trying to process
5897 the truncated data. It now detects and logs this case.
5899 17. Further to 4.21/58, another change has been made: if (and only if) the
5900 first line of a message (the first header line) ends with CRLF, a bare LF
5901 in a subsequent header line has a space inserted after it, so as not to
5902 terminate the header.
5904 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5905 unnecessarily, used atoi() instead of strtol(), and didn't check the
5906 termination when getting file sizes from file names by regex.
5908 19. Completely re-implemented the support for maildirsize files, in the light
5909 of a number of problems with the previous contributed implementation
5910 (4.30/29). In particular:
5912 . If the quota is zero, the maildirsize file is maintained, but no quota is
5915 . If the maildir directory does not exist, it is created before any attempt
5916 to write a maildirsize file.
5918 . The quota value in the file is just a cache; if the quota is changed in
5919 the transport, the new value overrides.
5921 . A regular expression is available for excluding directories from the
5924 20. The autoreply transport checks the characters in options that define the
5925 message's headers; it allows continued headers, but it was checking with
5926 isspace() after an embedded newline instead of explicitly looking for a
5929 21. If all the "regular" hosts to which an address was routed had passed their
5930 expiry times, and had not reached their retry times, the address was
5931 bounced, even if fallback hosts were defined. Now Exim should go on to try
5934 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5935 equivalent code in the SMTP transport. Some hosts send humungous responses
5936 to HELO/EHLO, more than 1024 it seems.
5938 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5939 gives compiler warnings in some environments. I've now done it "properly",
5942 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5943 (because of problems with the built-in one) was declared to return uschar *
5944 instead of char *, causing compiler failure.
5946 25. Fixed a file descriptor leak when processing alias/forward files.
5948 26. Fixed a minor format string issue in dbfn.c.
5950 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5952 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5953 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5956 29. When a sender address is verified, it is cached, to save repeating the test
5957 when there is more than one recipient in a message. However, when the
5958 verification involves a callout, it is possible for different callout
5959 options to be set for different recipients. It is too complicated to keep
5960 track of this in the cache, so now Exim always runs a verification when a
5961 callout is required, relying on the callout cache for the optimization.
5962 The overhead is duplication of the address routing, but this should not be
5965 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5966 to be verified with callout=postmaster, and the main callout worked but the
5967 postmaster check failed, the verification correctly failed. However, if a
5968 subsequent RCPT command asked for sender verification *without* the
5969 postmaster check, incorrect caching caused this verification also to fail,
5972 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5973 it was not caching the DNS options (qualify_single, search_parents) that
5974 were used when the lookup failed. A subsequent lookup with different
5975 options therefore always gave the same answer, though there were cases
5976 where it should not have. (Example: a "domains = !$mx_any" option on a
5977 dnslookup router: the "domains" option is always processed without any
5978 widening, but the router might have qualify_single set.) Now Exim uses the
5979 cached value only when the same options are set.
5981 32. Added John Jetmore's "exipick" utility to the distribution.
5983 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5984 than a timeout (e.g. a certificate is required, and is not provided), an
5985 Exim server now closes the connection immediately. Previously it waited for
5986 the client to close - but if the client is SSL, it seems that they each
5987 wait for each other, leading to a delay before one of them times out.
5989 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5990 maintained 0.8.x compatibility because I don't think many are using it, and
5991 it is clearly obsolete.
5993 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5996 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5997 client certificate was expired. A simple patch fixes this, though I don't
5998 understand the full logic of why the verify callback is called multiple
6001 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6002 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6003 which causes problems with some clients (such as the Certicom SSL Plus
6004 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6005 disables the coutermeasure allowing Eudora to connect."
6007 38. Exim was not checking that a write() to a log file succeeded. This could
6008 lead to Bad Things if a log got too big, in particular if it hit a file
6009 size limit. Exim now panics and dies if it cannot write to a log file, just
6010 as it does if it cannot open a log file.
6012 39. Modified OS/Makefile-Linux so that it now contains
6014 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6016 The two -D definitions ensure that Exim is compiled with large file
6017 support, which makes it possible to handle log files that are bigger than
6020 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6021 instance) a domain was checked against a named list that involved a lookup,
6022 causing $domain_data to be set, then another domain was checked against the
6023 same list, then the first domain was re-checked, the value of $domain_data
6024 after the final check could be wrong. In particular, if the second check
6025 failed, it could be set empty. This bug probably also applied to
6028 41. The strip_trailing_dot option was not being applied to the address given
6029 with the -f command-line option.
6031 42. The code for reading a message's header from the spool was incrementing
6032 $received_count, but never initializing it. This meant that the value was
6033 incorrect (doubled) while delivering a message in the same process in which
6034 it was received. In the most common configuration of Exim, this never
6035 happens - a fresh exec is done - but it can happen when
6036 deliver_drop_privilege is set.
6038 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6039 it now includes up to 150 characters of the unexpected data in the log
6042 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6043 and building data strings. The size of both of these buffers was 10 000
6044 bytes - far larger than anybody would *ever* want, thought I. Needless to
6045 say, somebody hit the limit. I have increased the maximum line length to
6046 20 000 and the maximum data length of concatenated lines to 100 000. I have
6047 also fixed two bugs, because there was no checking on these buffers. Tsk,
6048 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6049 buffer is too small.
6051 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6052 lsearch lookups. Now it does.
6054 46. When parsing a route_list item in a manualroute router, a fixed-length
6055 buffer was used for the list of hosts. I made this 1024 bytes long,
6056 thinking that nobody would ever have a list of hosts that long. Wrong.
6057 Somebody had a whole pile of complicated expansion conditions, and the
6058 string was silently truncated, leading to an expansion error. It turns out
6059 that it is easier to change to an unlimited length (owing to other changes
6060 that have happened since this code was originally written) than to build
6061 structure for giving a limitation error. The length of the item that
6062 expands into the list of hosts is now unlimited.
6064 47. The lsearch lookup could not handle data where the length of text line was
6065 more than 4095 characters. Such lines were truncated, leading to shortened
6066 data being returned. It should now handle lines of any length.
6068 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6069 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6072 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6073 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6074 no longer happens in comment lines. A list of replacements is now placed
6075 at the head of all of the source files, except those whose only change is
6076 to replace PERL_COMMAND in the very first #! line.
6078 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6079 messages on the queue, with a bottom-up merge sort, using code contributed
6080 by Michael Haardt. This should make operations like -bp somewhat faster on
6081 large queues. It won't affect queue runners, except when queue_run_in_order
6084 51. Installed eximstats 1.31 in the distribution.
6086 52. Added support for SRV lookups to the dnslookup router.
6088 53. If an ACL referred to $message_body or $message_body_end, the value was not
6089 reset for any messages that followed in the same SMTP session.
6091 54. The store-handling optimization for building very long strings was not
6092 differentiating between the different store pools. I don't think this
6093 actually made any difference in practice, but I've tidied it.
6095 55. While running the routers to verify a sender address, $sender_address
6096 was still set to the sender address. This is wrong, because when routing to
6097 send a bounce to the sender, it would be empty. Therefore, I have changed
6098 it so that, while verifying a sender address, $sender_address is set to <>.
6099 (There is no change to what happens when verifying a recipient address.)
6101 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6102 target A or AAAA records (if not already returned) without resetting the
6103 qualify_single or search_parents options of the DNS resolver. These are
6104 inappropriate in this case because the targets of MX and SRV records must
6105 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6106 target that, when qualified, matched something in the local domain. These
6107 two options are now turned off when doing these lookups.
6109 57. It seems that at least some releases of Reiserfs (which does not have the
6110 concept of a fixed number of inodes) returns zero and not -1 for the
6111 number of available inodes. This interacted badly with check_spool_inodes,
6112 which assumed that -1 was the "no such thing" setting. What I have done is
6113 to check that the total number of inodes is greater than zero before doing
6114 the test of how many are available.
6116 58. When a "warn" ACL statement has a log_message modifier, the message is
6117 remembered, and not repeated. This is to avoid a lot of repetition when a
6118 message has many recipients that cause the same warning to be written.
6119 However, Exim was preserving the list of already written lines for an
6120 entire SMTP session, which doesn't seem right. The memory is now reset if a
6121 new message is started.
6123 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6124 debugging output unless log_rewrite was also set.
6126 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6127 of (int)(handle) when we know that handle contains (void *)(-1).
6129 61. The Exim daemon panic-logs an error return when it closes the incoming
6130 connection. However "connection reset by peer" seems to be common, and
6131 isn't really an error worthy of noting specially, so that particular error
6134 62. When Exim is trying to find all the local interfaces, it used to panic and
6135 die if the ioctl to get the interface flags failed. However, it seems that
6136 on at least one OS (Solaris 9) it is possible to have an interface that is
6137 included in the list of interfaces, but for which you get a failure error
6138 for this call. This happens when the interface is not "plumbed" into a
6139 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6140 failure of the "get flags" call assumes that the interface is down.
6142 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6143 makes life easier for people who are doing arithmetic on fields extracted
6144 from dates, where you often get leading zeros that should not be
6145 interpreted as octal.
6147 64. Added qualify_domain to the redirect router, to override the global
6150 65. If a pathologically long header line contained very many addresses (the
6151 report of this problem mentioned 10 000) and each of them was rewritten,
6152 Exim could use up a very large amount of memory. (It kept on making new
6153 copies of the header line as it rewrote, and never released the old ones.)
6154 At the expense of a bit more processing, the header rewriting function has
6155 been changed so that it no longer eats memory in this way.
6157 66. The generation of the Received: header has been moved from the time that a
6158 message starts to be received, to the time that it finishes. The timestamp
6159 in the Received: header should now be very close to that of the <= log
6160 line. There are two side-effects of this change:
6162 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6163 logged header lines no longer include the local Received: line, because
6164 it has not yet been created. The same applies to a copy of the message
6165 that is returned to a non-SMTP sender when a message is rejected.
6167 (b) When a filter file is tested using -bf, no additional Received: header
6168 is added to the test message. After some thought, I decided that this
6171 This change does not affect the value of $received_for. It is still set
6172 after address rewriting, but before local_scan() is called.
6174 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6176 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6177 gave an unhelpful panic error message, and a defer error. I have managed to
6178 change this behaviour so that it now rejects any supplied certificate,
6179 which seems right, as the list of acceptable certificates is empty.
6181 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6182 gave an unhelpful defer error. I have not managed to make this reject any
6183 supplied certificates, but the error message it gives is "no certificate
6184 supplied", which is not helpful.
6186 70. exigrep's output now also includes lines that are not associated with any
6187 message, but which match the given pattern. Implemented by a patch from
6188 Martin Sluka, which also tidied up the Perl a bit.
6190 71. Recipient callout verification, like sender verification, was using <> in
6191 the MAIL FROM command. This isn't really the right thing, since the actual
6192 sender may affect whether the remote host accepts the recipient or not. I
6193 have changed it to use the actual sender in the callout; this means that
6194 the cache record is now keyed on a recipient/sender pair, not just the
6195 recipient address. There doesn't seem to be a real danger of callout loops,
6196 since a callout by the remote host to check the sender would use <>.
6197 [SEE ABOVE: changed after hitting problems.]
6199 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6200 temporary errors. However, in the case of such a code being given after
6201 the end of a data transmission (i.e. after ".") Exim was failing to write
6202 a retry record for the message. (Yes, there was some broken host that was
6203 actually sending 8xx at this point.)
6205 73. An unknown lookup type in a host list could cause Exim to panic-die when
6206 the list was checked. (An example that provoked this was putting <; in the
6207 middle of a list instead of at the start.) If this happened during a DATA
6208 ACL check, a -D file could be left lying around. This kind of configuration
6209 error no longer causes Exim to die; instead it causes a defer error. The
6210 incident is still logged to the main and panic logs.
6212 74. Buglet left over from Exim 3 conversion. The message "too many messages
6213 in one connection" was written to the rejectlog but not the mainlog, except
6214 when address rewriting (yes!) was being logged.
6216 75. Added write_rejectlog option.
6218 76. When a system filter was run not as root (that is, when system_filter_user
6219 was set), the values of the $n variables were not being returned to the
6220 main process; thus, they were not subsequently available in the $sn
6223 77. Added +return_path_on_delivery log selector.
6225 78. A connection timeout was being treated differently from recipients deferred
6226 when testing hosts_max_try with a message that was older than the host's
6227 retry timeout. (The host should not be counted, thus allowing all hosts to
6228 be tried at least once before bouncing.) This may have been the cause of an
6229 occasionally reported bug whereby a message would remain on the queue
6230 longer than the retry timeout, but would be bounced if a delivery was
6231 forced. I say "may" because I never totally pinned down the problem;
6232 setting up timeout/retry tests is difficult. See also the next item.
6234 79. The ultimate address timeout was not being applied to errors that involved
6235 a combination of host plus message (for example, a timeout on a MAIL
6236 command). When an address resolved to a number of possible hosts, and they
6237 were not all tried for each delivery (e.g. because of hosts_max_try), a
6238 message could remain on the queue longer than the retry timeout.
6240 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6243 81. Fixed an obscure SMTP outgoing bug which required at least the following
6244 conditions: (a) there was another message waiting for the same server;
6245 (b) the server returned 5xx to all RCPT commands in the first message so
6246 that the message was not completed; (c) the server dropped the connection
6247 or gave a negative response to the RSET that Exim sends to abort the
6248 transaction. The observed case was a dropped connection after DATA that had
6249 been sent in pipelining mode. That is, the server had advertised PIPELINING
6250 but was not implementing it correctly. The effect of the bug was incorrect
6251 behaviour, such as trying another host, and this could lead to a crash.
6257 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6258 and daemon.c were passed as pointers to ints; they should have been
6259 pointers to socklen_t variables (which are typically unsigned ints).
6261 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6264 3. Fixed a really odd bug that affected only the testing scheme; patching a
6265 certain fixed string in the binary changed the value of another string that
6266 happened to be identical to the end of the original first string.
6268 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6269 name", it returns that address as the IP address. On some operating
6270 systems (e.g. Solaris), it also passes back the IP address string as the
6271 "host name". However, on others (e.g. Linux), it passes back an empty
6272 string. Exim wasn't checking for this, and was changing the host name to an
6273 empty string, assuming it had been canonicalized.
6275 5. Although rare, it is permitted to have more than one PTR record for a given
6276 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6277 all the names associated with an address, because they do in Solaris.
6278 However, it seems that they do not in Linux for data that comes from the
6279 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6280 I found this out when I moved to a new Linux workstation and tried to run
6281 the Exim test suite.
6283 To get round this problem I have changed the code so that it now does its
6284 own call to the DNS to look up PTR records when searching for a host name.
6285 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6286 addresses that are only in /etc/hosts are still found.
6288 This behaviour is, however, controlled by an option called host_lookup_
6289 order, which defaults to "bydns:byaddr". If people want to use the other
6290 order, or indeed, just use one or the other means of lookup, they can
6291 specify it in this variable.
6293 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6294 some operating systems, this comes back from gethostbyaddr() as an empty
6295 string, and this is what Exim used to test for. However, it seems that in
6296 other systems, "." is yielded. Exim now tests for this case too.
6298 7. The values of check_spool_space and check_log_space are now held internally
6299 as a number of kilobytes instead of an absolute number of bytes. If a
6300 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6301 kilobyte. This means that much larger values can be stored.
6303 8. Exim monitor: an attempt to get the action menu when not actually pointing
6304 at a message produces an empty menu entitled "No message selected". This
6305 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6306 no entries in it ("Shell widget menu has zero width and/or height"). So I
6307 have added a single, blank menu entry in this case.
6309 9. Added ${quote_local_part.
6311 10. MIME decoding is now applied to the contents of Subject: header lines when
6314 11. Now that a reference to $sender_host_address automatically causes a reverse
6315 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6316 host lookup before query-style lookups in lists that might use this
6317 variable. This has therefore been abolished, and the "net-" prefix is no
6318 longer necessary for query-style lookups.
6320 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6321 have been a typo for LFLAGS, so it has been changed.
6323 13. The install script calls Exim with "-C /dev/null" in order to find the
6324 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6325 to be output. However, since Exim outputs its version number before the
6326 error, it didn't break the script. It just looked ugly. I fixed this by
6327 always allowing "-C /dev/null" if the caller is root.
6329 14. Ignore overlarge ACL variable number when reading spool file - insurance
6330 against a later release with more variables having written the file.
6332 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6333 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6334 this, as well as the form without the "IPv6" on the front.
6336 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6337 OS/Makefile-Darwin file.
6339 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6340 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6342 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6343 variable. However, some OS use ints for inodes. I've added cast to long int
6344 to get rid of the compiler warning.
6346 19. I had forgotten to lock out "/../" in configuration file names when
6347 ALT_CONFIG_PREFIX was set.
6349 20. Routers used for verification do not need to specify transports. However,
6350 if such a router generated a host list, and callout was configured, Exim
6351 crashed, because it could not find a port number from the (non-existent)
6352 transport. It now assumes port 25 in this circumstance.
6354 21. Added the -t option to exigrep.
6356 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6357 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6358 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6359 them is set, LOOKUP_LSEARCH is forced.
6361 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6363 24. Added sender and host information to the "rejected by local_scan()" log
6364 line; previously there was no indication of these.
6366 25. Added .include_if_exists.
6368 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6369 Linux. It turns out that not all file systems support this. Apparently some
6370 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6371 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6372 not supported on the file descriptor, is now ignored when Exim is trying to
6373 sync a directory. This applies only to Linux.
6375 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6377 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6378 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6381 29. Added support for maildirsize files from supplied patch (modified a bit).
6383 30. The use of :fail: followed by an empty string could lead Exim to respond to
6384 sender verification failures with (e.g.):
6386 550 Verification failed for <xxx>
6387 550 Sender verify failed
6389 where the first response line was missing the '-' that indicates it is not
6390 the final line of the response.
6392 31. The loop for finding the name of the user that called Exim had a hardwired
6393 limit of 10; it now uses the value of finduser_retries, which is used for
6394 all other user lookups.
6396 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6399 33. Exim was neglecting to zero errno before one call of strtol() when
6400 expanding a string and expecting an integer value. On some systems this
6401 resulted in spurious "integer overflow" errors. Also, it was casting the
6402 result into an int without checking.
6404 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6405 did not work. The code looks as if it has *never* worked, though it appears
6406 to have been documented since at least release 1.62. I have made it work.
6408 35. The "timeout_DNS" error in retry rules, also documented since at least
6409 1.62, also never worked. As it isn't clear exactly what this means, and
6410 clearly it isn't a major issue, I have abolished the feature by treating it
6411 as "timeout", and writing a warning to the main and panic logs.
6413 36. The display of retry rules for -brt wasn't always showing the error code
6416 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6417 timeout_connect_A, timeout_connect_MX.
6419 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6420 to the empty sender.
6422 39. The daemon was not analysing the content of -oX till after it had closed
6423 stderr and disconnected from the controlling terminal. This meant that any
6424 syntax errors were only noted on the panic log, and the return code from
6425 the command was 0. By re-arranging the code a little, I've made the
6426 decoding happen first, so such errors now appear on stderr, and the return
6427 code is 1. However, the actual setting up of the sockets still happens in
6428 the disconnected process, so errors there are still only recorded on the
6431 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6432 connections (as happens on some IP stacks) was logged at start up time as
6433 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6434 it from "IPv6 and IPv4", which means that two separate sockets are being
6437 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6438 says whether AF_INET or AF_INET6 was passed as an argument.
6440 42. Exiwhat output was messed up when time zones were included in log
6443 43. Exiwhat now gives more information about the daemon's listening ports,
6444 and whether -tls-on-connect was used.
6446 44. The "port" option of the smtp transport is now expanded.
6448 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6449 silently ignored. Now an error message is written to the main and panic
6452 46. There's a new ACL modifier called "logwrite" which writes to a log file
6453 as soon as it is encountered.
6455 47. Added $local_user_uid and $local_user_gid at routing time.
6457 48. Exim crashed when trying to verify a sender address that was being
6460 49. Exim was recognizing only a space character after ".include". It now also
6461 recognizes a tab character.
6463 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6464 extracting the relevant information from the specification. The man page no
6465 longer contains scrambled data for the -d option, and I've added a section
6466 at the front about calling Exim under different names.
6468 51. Added "extra_headers" argument to the "mail" command in filter files.
6470 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6473 53. Installed eximstats 1.29.
6475 54. Added transport_filter_timeout as a generic transport option.
6477 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6478 Cc: header lines. This was required by RFC 822, but it not required by RFC
6481 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6482 incoming messages that did not have them. Now it does so only if the
6483 message originates locally, that is, if there is no associated remote host
6484 address. When Resent- header lines are present, this applies to the Resent-
6485 lines rather than the non-Resent- lines.
6487 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6488 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6490 58. Messages for configuration errors now include the name of the main
6491 configuration file - useful now that there may be more than one file in a
6492 list (.included file names were always shown).
6494 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6495 for those rare installations that do not start the daemon as root or run it
6496 setuid root. I've cut out the call to initgroups() if the daemon is not
6499 60. The Exim user and group can now be bound into the binary as text strings
6500 that are looked up at the start of Exim's processing.
6502 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6504 62. Added $mailstore_basename variable.
6506 63. Installed patch to sieve.c from Michael Haardt.
6508 64. When Exim failed to open the panic log after failing to open the main log,
6509 the original message it was trying to log was written to stderr and debug
6510 output, but if they were not available (the usual case in production), it
6511 was lost. Now it is written to syslog before the two lines that record the
6512 failures to open the logs.
6514 65. Users' Exim filters run in subprocesses under the user's uid. It is
6515 possible for a "deliver" command or an alias in a "personal" command to
6516 provoke an address rewrite. If logging of address rewriting is configured,
6517 this fails because the process is not running as root or exim. There may be
6518 a better way of dealing with this, but for the moment (because 4.30 needs
6519 to be released), I have disabled address rewrite logging when running a
6520 filter in a non-root, non-exim process.
6526 1. The buildconfig auxiliary program wasn't quoting the value set for
6527 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6528 not defined. This bug was masked in 4.22 by the effect that was fixed in
6531 2. Some messages that were rejected after a message id was allocated were
6532 shown as "incomplete" by exigrep. It no longer does this for messages that
6533 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6535 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6536 have allow_domain_literals set, the ID did not get logged in the <= line.
6537 Domain literals are now always recognized in Message-ID: header lines.
6539 4. The first argument for a ${extract expansion item is the key name or field
6540 number. Leading and trailing spaces in this item were not being ignored,
6541 causing some misleading effects.
6543 5. When deliver_drop_privilege was set, single queue runner processes started
6544 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6545 same command in the process it spins off) were not dropping privilege.
6547 6. When the daemon running as "exim" started a queue runner, it always
6548 re-executed Exim in the spun-off process. This is a waste of effort when
6549 deliver_drop_privilege is set. The new process now just calls the
6550 queue-runner function directly.
6556 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6559 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6560 ignored. Though the use of -oP was forcing the writing of a pid file, it
6561 was always written to the default place.
6563 3. If the message "no IP address found for host xxxx" is generated during
6564 incoming verification, it is now followed by identification of the incoming
6565 connection (so you can more easily find what provoked it).
6567 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6569 5. Added some features to "harden" Exim a bit more against certain attacks:
6571 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6572 be put in Local/Makefile. This is like the never_users runtime option,
6573 but it cannot be overridden. The default setting is "root".
6575 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6576 prefix string with which any file named in a -C command line option
6579 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6580 is retained for -C and -D only if the caller of Exim is root. Without
6581 it, the exim user may also use -C and -D and retain privilege.
6583 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6584 command line option is disabled.
6586 6. Macro names set by the -D option must start with an upper case letter, just
6587 like macro names defined in the configuration file.
6589 7. Added "dereference=" facility to LDAP.
6591 8. Two instances of the typo "uknown" in the source files are fixed.
6593 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6594 the Configure-Makefile script screwed up while processing it.
6596 10. Incorporated PCRE 4.4.
6598 11. The SMTP synchronization check was not operating right at the start of an
6599 SMTP session. For example, it could not catch a HELO sent before the client
6600 waited for the greeting. There is now a check for outstanding input at the
6601 point when the greeting is written. Because of the duplex, asynchronous
6602 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6603 way, but not yet received, when the check is performed.
6605 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6606 on TCP/IP sockets, because this apparently causes some broken clients to
6609 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6610 unchanged) from the Cygwin maintainer.
6612 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6613 was supported instead of testing for mbx. Effectively a typo.
6615 15. The spa authenticator server code was not checking that the input it
6616 received was valid base64.
6618 16. The debug output line for the "set" modifier in ACLs was not showing the
6619 name of the variable that was being set.
6621 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6623 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6624 lookup on its own. Something else was needed to trigger the lookup. For
6625 example, a match in host_lookup or the need for a host name in a host list.
6626 Now, if $sender_host_name is referenced and the host name has not yet been
6627 looked up, a lookup is performed. If the lookup fails, the variable remains
6628 empty, and $host_lookup_failed is set to "1".
6630 19. Added "eqi" as a case-independent comparison operator.
6632 20. The saslauthd authentication condition could segfault if neither service
6633 nor realm was specified.
6635 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6636 error message that was logged was misleading, and incoming SMTP
6637 connections were dropped. The message is now more accurate, and temporary
6638 errors are given to SMTP connections.
6640 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6641 (except RSET) with a 421 error, until QUIT is received. However, it was
6642 failing to send a response to QUIT.
6644 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6645 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6646 ACL is now run after the helo_try_verify_hosts code.
6648 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6649 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6650 case-independent, so other case variants are also recognized). Apparently
6651 some systems use these upper case variants.
6653 25. If more than two messages were waiting for the same host, and a transport
6654 filter was specified for the transport, Exim sent two messages over the
6655 same TCP/IP connection, and then failed with "socket operation on non-
6656 socket" when it tried to send the third.
6658 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6660 27. The extern definition of crypt16() in expand.c was not being excluded when
6661 the OS had its own crypt16() function.
6663 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6664 as a preferred synonym for return_size_limit, both as an option and as an
6667 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6669 30. Changed the default configuration ACL to relax the local part checking rule
6670 for addresses that are not in any local domains. For these addresses,
6671 slashes and pipe symbols are allowed within local parts, but the sequence
6672 /../ is explicitly forbidden.
6674 31. SPA server authentication was not clearing the challenge buffer before
6677 32. log_message in a "warn" ACL statement was writing to the reject log as
6678 well as to the main log, which contradicts the documentation and doesn't
6679 seem right (because no rejection is happening). So I have stopped it.
6681 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6682 However, I am unable to do any testing of this.
6684 34. Fixed an infelicity in the appendfile transport. When checking directories
6685 for a mailbox, to see if any needed to be created, it was accidentally
6686 using path names with one or more superfluous leading slashes; tracing
6687 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6689 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6690 contents are added to the log line that is written for every discarded
6691 recipient. (Previously a log_message setting was ignored.)
6693 36. The ${quote: operator now quotes the string if it is empty.
6695 37. The install script runs exim in order to find its version number. If for
6696 some reason other than non-existence or emptiness, which it checks, it
6697 could not run './exim', it was installing it with an empty version number,
6698 i.e. as "exim-". This error state is now caught, and the installation is
6701 38. An argument was missing from the function that creates an error message
6702 when Exim fails to connect to the socket for saslauthd authentication.
6703 This could cause Exim to crash, or give a corrupted message.
6705 39. Added isip, isip4, and isip6 to ${if conditions.
6707 40. The ACL variables $acl_xx are now saved with the message, and can be
6708 accessed later in routers, transports, and filters.
6710 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6711 strings in the file are not string-expanded.
6713 42. If a MAIL command specified a SIZE value that was too large to fit into an
6714 int variable, the check against message_size_limit failed. Such values are
6715 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6716 day this will have to be increased, but I don't think I want to be around
6717 when emails are that large.
6724 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6725 iconv() is not standard in FreeBSD.
6727 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6728 IPv6 enabled. The observed symptom was a segmentation fault on return from
6729 the function os_common_find_running_interfaces() in src/os.c.
6731 3. In the check_special_case() function in daemon.c I had used "errno" as an
6732 argument name, which causes warnings on some systems. This was basically a
6733 typo, since it was named "eno" in the comments!
6735 4. The code that waits for the clock to tick (at a resolution of some fraction
6736 of a second) so as to ensure message-id uniqueness was always waiting for
6737 at least one whole tick, when it could have waited for less. [This is
6738 almost certainly not relevant at current processor speeds, where it is
6739 unlikely to ever wait at all. But we try to future-proof.]
6741 5. The function that sleeps for a time interval that includes fractions of a
6742 second contained a race. It did not block SIGALRM between setting the
6743 timer, and suspending (a couple of lines later). If the interval was short
6744 and the sigsuspend() was delayed until after it had expired, the suspension
6745 never ended. On busy systems this could lead to processes getting stuck for
6748 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6749 process, before it forks any delivery processes. The open lookup caching
6750 mechanism meant that the open file or database connection was passed into
6751 the delivery process. The problem was that delivery processes always tidy
6752 up cached lookup data. This could cause a problem for the next delivery
6753 process started by the queue runner, because the external queue runner
6754 process does not know about the closure. So the next delivery process
6755 still has data in the lookup cache. In the case of a file lookup, there was
6756 no problem because closing a file descriptor in a subprocess doesn't affect
6757 the parent. However, if the lookup was caching a connection to a database,
6758 the connection was closed, and the second delivery process was likely to
6759 see errors such as "PGSQL: query failed: server closed the connection
6760 unexpectedly". The problem has been fixed by closing all cached lookups
6761 in a queue runner before running a delivery process.
6763 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6764 seem to have the "const" qualifier which it has on other OS. I've
6767 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6768 *of the same type* (client or server) with the same public name that an
6769 error should be diagnosed.
6771 9. When Exim looked up a host name for an IP address, but failed to find the
6772 original IP address when looking up the host name (a safety check), it
6773 output the message "<ip address> does not match any IP for NULL", which was
6774 confusing, to say the least. The bug was that the host name should have
6775 appeared instead of "NULL".
6777 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6778 user that is built into the binary, and the -C or -D options is used, root
6779 privilege is dropped before the configuration file is read. In addition,
6780 logging is switched to stderr instead of the normal log files. If the
6781 configuration then re-defines the Exim user, the unprivileged environment
6782 is probably not what is expected, so Exim logs a panic warning message (but
6785 However, if deliver_drop_privilege is set, the unprivileged state may well
6786 be exactly what is intended, so the warning has been cut out in that case,
6787 and Exim is allowed to try to write to its normal log files.
6793 1. smtp_return_error_details was not giving details for temporary sender
6794 or receiver verification errors.
6796 2. Diagnose a configuration error if two authenticators have the same public
6799 3. Exim used not to create the message log file for a message until the first
6800 delivery attempt. This could be confusing when incoming messages were held
6801 for policy or load reasons. The message log file is now created at the time
6802 the message is received, and an initial "Received" line is written to it.
6804 4. The automatically generated man page for command line options had a minor
6805 bug that caused no ill effects; however, a more serious problem was that
6806 the procedure for building the man page automatically didn't always
6807 operate. Consequently, release 4.20 contains an out-of-date version. This
6808 shouldn't happen again.
6810 5. When building Exim with embedded Perl support, the script that builds the
6811 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6812 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6814 6. The freeze_tell option was not being used for messages that were frozen on
6815 arrival, either by an ACL or by local_scan().
6817 7. Added the smtp_incomplete_transaction log selector.
6819 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6820 was accepting AUTH without a new EHLO.
6822 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6823 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6826 10. Logging of TCP/IP connections (when configured) now happens in the main
6827 daemon process instead of the child process, so that the TCP/IP connection
6828 count is more accurate (but it can never be perfect).
6830 11. The use of "drop" in a nested ACL was not being handled correctly in the
6831 outer ACL. Now, if condition failure induced by the nested "drop" causes
6832 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6833 or "require"), the connection is dropped.
6835 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6836 that yield "discard" can now be used with an "accept" or a "discard" verb,
6837 but an error is generated for any others (because I can't see a useful way
6838 to define what should happen).
6840 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6841 are now processed in the same way as lines in the Exim configuration file.
6842 In particular, continuation lines are supported.
6844 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6846 15. Added -ti meaning -t -i.
6848 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6849 domains, and warn by logging if others are found.
6851 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6852 returned from GIFCONF when Exim is trying to find the list of interfaces on
6853 a host. The code in os.c has been modified to copy each ifreq to an aligned
6854 structure in all cases.
6856 Also, in some cases, the returned ifreq's were being copied to a 'struct
6857 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6858 means the last couple of bytes of an IPv6 address could be chopped if the
6859 ifreq contained only a normal sockaddr (14 bytes storage).
6861 18. Named domain lists were not supported in the hosts_treat_as_local option.
6862 An entry such as +xxxx was not recognized, and was treated as a literal
6865 19. Ensure that header lines added by a DATA ACL are included in the reject log
6866 if the ACL subsequently rejects the message.
6868 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6869 MD5 (which is deprecated).
6871 21. When testing a filter file using -bf, Exim was writing a message when it
6872 took the sender from a "From " line in the message, but it was not doing so
6873 when it took $return_path from a Return-Path: header line. It now does.
6875 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6876 with a valid header line field name (a series of printing characters
6877 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6879 23. Changed "disc" in the source to "disk" to conform to the documentation and
6880 the book and for uniformity.
6882 24. Ignore Sendmail's -Ooption=value command line item.
6884 25. When execve() failed while trying to run a command in a pipe transport,
6885 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6886 could be confused with a return value of 69 from the command itself. This
6887 has been changed to 127, the value the shell returns if it is asked to run
6888 a non-existent command. The wording for the related log line suggests a
6889 non-existent command as the problem.
6891 26. If received_header_text expands to an empty string, do not add a Received:
6892 header line to the message. (Well, it adds a token one on the spool, but
6893 marks it "old" so that it doesn't get used or transmitted.)
6895 27. Installed eximstats 1.28 (addition of -nt option).
6897 28. There was no check for failure on the call to getsockname() in the daemon
6898 code. This can fail if there is a shortage of resources on the system, with
6899 ENOMEM, for example. A temporary error is now given on failure.
6901 29. Contrary to the C standard, it seems that in some environments, the
6902 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6903 program. Exim now does this explicitly; it affects the formatting of
6904 timestamps using strftime().
6906 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6907 complaints. I've now initialized all the variables, to avoid this.
6909 32. Header lines added by a system filter were not being "seen" during
6910 transport-time rewrites.
6912 33. The info_callback() function passed to OpenSSL is set up with type void
6913 (*)(SSL *, int, int), as described somewhere. However, when calling the
6914 function (actually a macro) that sets it up, the type void(*)() is
6915 expected. I've put in a cast to prevent warnings from picky compilers.
6917 34. If a DNS black list lookup found a CNAME record, but there were no A
6918 records associated with the domain it pointed at, Exim crashed.
6920 35. If a DNS black list lookup returned more than one A record, Exim ignored
6921 all but the first. It now scans all returned addresses if a particular IP
6922 value is being sought. In this situation, the contents of the
6923 $dnslist_value variable are a list of all the addresses, separated by a
6926 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6927 to accept a host name and all its aliases if the forward lookup for any of
6928 them yielded the IP address of the incoming connection. Now it accepts only
6929 those names whose forward lookup yields the correct IP address. Any other
6930 names are discarded. This closes a loophole whereby a rogue DNS
6931 administrator could create reverse DNS records to break through a
6932 wildcarded host restriction in an ACL.
6934 37. If a user filter or a system filter that ran in a subprocess used any of
6935 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6936 the wrong values were passed to the pipe command ($thisaddress had the
6937 value of $0, $0 had the value of $1, etc). This bug was introduced by
6938 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6940 38. Improved the line breaking for long SMTP error messages from ACLs.
6941 Previously, if there was no break point between 40 and 75 characters, Exim
6942 left the rest of the message alone. Two changes have been made: (a) I've
6943 reduced the minimum length to 35 characters; (b) if it can't find a break
6944 point between 35 and 75 characters, it looks ahead and uses the first one
6945 that it finds. This may give the occasional overlong line, but at least the
6946 remaining text gets split now.
6948 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6949 file descriptors might be low, and that setting 1000 would always raise it.
6950 It turns out that in some environments, the limit is already over 1000 and
6951 that lowering it causes trouble. So now Exim takes care not to decrease it.
6953 40. When delivering a message, the value of $return_path is set to $sender_
6954 address at the start of routing (routers may change the value). By an
6955 oversight, this default was not being set up when an address was tested by
6956 -bt or -bv, which affected the outcome if any router or filter referred to
6959 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6960 log only when "message" or "log_message" are set. However, if one of the
6961 conditions was an address verification, or a call to a nested ACL, the
6962 messages generated by the underlying test were being passed through. This
6963 no longer happens. The underlying message is available in $acl_verify_
6964 message for both "message" and "log_message" expansions, so it can be
6965 passed through if needed.
6967 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6968 new expansion $bh_ to give the encoded byte string without charset
6969 translation. Translation happens only if iconv() is available; HAVE_ICONV
6970 indicates this at build time. HEADERS_CHARSET gives the charset to
6971 translate to; headers_charset can change it in the configuration, and
6972 "headers charset" can change it in an individual filter file.
6974 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6975 that creates RFC 2047 encoded "words" labels them as that charset instead
6976 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6977 expansion operator; (ii) when Exim creates a From: line for a local
6978 message; (iii) when a header line is rewritten to include a "phrase" part.
6980 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6981 buggy, causing it to skip the first lines of messages whose message ID
6982 ended in 'D'. This would not have bitten before Exim release 4.14, because
6983 message IDs were unlikely to end in 'D' before then. The effect was to have
6984 incorrect size information for certain domains.
6986 45. #include "config.h" was missing at the start of the crypt16.c module. This
6987 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6990 46. If there was a timeout during a "random" callout check, Exim treated it as
6991 a failure of the random address, and carried on sending RSET and the real
6992 address. If the delay was just some slowness somewhere, the response to the
6993 original RCPT would be taken as a response to RSET and so on, causing
6994 mayhem of various kinds.
6996 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6997 when I implemented it. It didn't allow for the fact that some option values
6998 may legitimately be negative (e.g. size_addition), and it didn't even do
6999 the right test for positive values.
7001 48. Domain names in DNS records are case-independent. Exim always looks them up
7002 in lower case. Some resolvers return domain names in exactly the case they
7003 appear in the zone file, that is, they may contain uppercase letters. Not
7004 all resolvers do this - some return always lower case. Exim was treating a
7005 change of case by a resolver as a change of domain, similar to a widening
7006 of a domain abbreviation. This triggered its re-routing code and so it was
7007 trying to route what was effectively the same domain again. This normally
7008 caused routing to fail (because the router wouldn't handle the domain
7009 twice). Now Exim checks for this case specially, and just changes the
7010 casing of the domain that it ultimately uses when it transmits the message
7013 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7016 50. If a filter generated a file delivery with a non-absolute name (possible if
7017 no home directory exists for the router), the forbid_file option was not
7020 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7021 the existing equality matching.
7023 52. Exim was using ints instead of ino_t variables in some places where it was
7024 dealing with inode numbers.
7026 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7027 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7028 called TMPDIR, and if it finds it is different, it changes its value.
7030 54. The smtp_printf() function is now made available to local_scan() so
7031 additional output lines can be written before returning. There is also an
7032 smtp_fflush() function to enable the detection of a dropped connection.
7033 The variables smtp_input and smtp_batched_input are exported to
7036 55. Changed the default runtime configuration: the message "Unknown user"
7037 has been removed from the ACL, and instead placed on the localuser router,
7038 using the cannot_route_message feature. This means that any verification
7039 failures that generate their own messages won't get overridden. Similarly,
7040 the "Unrouteable address" message that was in the ACL for unverifiable
7041 relay addresses has also been removed.
7043 56. Added hosts_avoid_esmtp to the smtp transport.
7045 57. The exicyclog script was not checking for the esoteric option
7046 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7047 will work only if exicyclog is run under the appropriate euid.
7049 58. Following a discussion on the list, the rules by which Exim recognises line
7050 endings on incoming messages have been changed. The -dropcr and drop_cr
7051 options are now no-ops, retained only for backwards compatibility. The
7052 following line terminators are recognized: LF CRLF CR. However, special
7053 processing applies to CR:
7055 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7056 nor a local message in the state where . is a terminator.
7058 (ii) If a bare CR is encountered in a header line, an extra space is added
7059 after the line terminator so as not to end the header. The reasoning
7060 behind this is that bare CRs in header lines are most likely either
7061 to be mistakes, or people trying to play silly games.
7063 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7064 was being incorrectly given as 18 bytes larger than it should have been.
7065 This is a VOB (very old bug).
7067 60. This may never have affected anything current, but just in case it has:
7068 When the local host is found other than at the start of a list of hosts,
7069 the local host, those with the same MX, and any that follow, are discarded.
7070 When the list in question was part of a longer list of hosts, the following
7071 hosts (not currently being processed) were also being discarded. This no
7072 longer happens. I'm not sure if this situation could ever has previously
7075 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7078 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7079 header. This is something that is recommended in a new Internet Draft, and
7080 is something that is documented as being done by Sendmail. There are two
7081 possible values. For messages generated by the autoreply transport, Exim
7084 Auto-Submitted: auto-replied
7086 whereas for all other generated messages (e.g. bounces) it adds
7088 Auto-Submitted: auto-generated
7090 63. The "personal" condition in filters now includes a test for the
7091 Auto-Submitted: header. If it contains the string "auto-" the message it
7092 not considered personal.
7094 64. Added rcpt_include_affixes as a generic transport option.
7096 65. Added queue_only_override (default true).
7098 66. Added the syslog_duplication option.
7100 67. If what should have been the first header line of a message consisted of
7101 a space followed by a colon, Exim was mis-interpreting it as a header line.
7102 It isn't of course - it is syntactically invalid and should therefore be
7103 treated as the start of the message body. The misbehaviour could have
7104 caused a number of strange effects, including loss of data in subsequent
7105 header lines, and spool format errors.
7107 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7108 client host had authenticated. This control can now be exercised by an ACL
7109 for more flexibility.
7111 69. By default, callouts do not happen when testing with -bh. There is now a
7112 variant, -bhc, which does actually run the callout code, including
7113 consulting and updating the callout cache.
7115 70. Added support for saslauthd authentication, courtesy of Alexander
7118 71. If statvfs() failed on the spool or log directories while checking their
7119 size for availability, Exim confusingly gave the error "space shortage".
7120 Furthermore, in debugging mode it crashed with a floating point exception.
7121 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7122 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7123 serious problem, Exim now writes to the main and panic logs when this
7124 happens, with details of the failure. It then refuses to accept the
7125 incoming message, giving the message "spool directory problem" or "log
7126 directory problem" with a 421 code for SMTP messages.
7128 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7129 0, 1, and 2 exist, because some OS complain for execs without them (see
7130 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7131 descriptors. However, the code omitted to check that the open succeeded,
7132 causing mysterious errors if for some reason the permissions on /dev/null
7133 got screwed. Now Exim writes a message to the main and panic logs, and
7134 bombs out if it can't open /dev/null.
7136 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7137 interact so that it is all more flexible. It is supposed to remain
7138 backwards compatible. Also added extra_local_interfaces.
7140 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7141 to bomb out with an assertion failure - to the client this appears as a
7142 connection drop. This problem occurs in the part of the code that was taken
7143 from the Samba project. Fortunately, the assertion is in a very simple
7144 function, so I have fixed this by reproducing the function inline in the
7145 one place where it is called, and arranging for authentication to fail
7146 instead of killing the process with assert().
7148 75. The SPA client code was not working when the server requested OEM rather
7149 than Unicode encoding.
7151 76. Added code to make require_files with a specific uid setting more usable in
7152 the case where statting the file as root fails - usually a non-root-mounted
7153 NFS file system. When this happens and the failure is EACCES, Exim now
7154 forks a subprocess and does the per-uid checking as the relevant uid.
7156 77. Added process_log_path.
7158 78. If log_file_path was not explicitly set, a setting of check_log_space or
7159 check_log_inodes was ignored.
7161 79. If a space check for the spool or log partitions fails, the incident is now
7162 logged. Of course, in the latter case the data may get lost...
7164 80. Added the %p formatting code to string_format() so that it can be used to
7165 print addresses in debug_print(). Adjusted all the address printing in the
7166 debugging in store.c to use %p rather than %d.
7168 81. There was a concern that a line of code in smtp_in.c could overflow a
7169 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7170 initially expressed, the concern was not well-founded, because trailing
7171 spaces are removed early. However, if the trailing spaces were followed by
7172 a NULL, they did not get removed, so the overflow was possible. Two fixes
7175 (a) I re-wrote the offending code in a cleaner fashion.
7176 (b) If an incoming SMTP command contains a NULL character, it is rejected
7179 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7180 runs initgroups(), so that if the Exim user is in any additional groups,
7181 they will be used during message reception.
7187 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob