1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
31 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
34 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
35 Previously this would segfault.
37 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
40 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
42 JH/16 Fix ARC verification to do AS checks in reverse order.
44 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
45 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
46 Previously the "utf8" would be re-prepended for every additional message.
48 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
49 one parent address was copied, and bogus data was used at delivery-logging
50 time. Either a crash (after delivery) or bogus log data could result.
51 Discovery and analysis by Tim Stewart.
53 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
54 The ignore_error flag wasn't being returned from the filter subprocess so
55 was not set for later routers. Investigation and fix by Matthias Kurz.
57 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
58 requiring a known-CA anchor certificate; make it now rely entirely on the
59 TLSA as an anchor. Checking the name on the leaf cert against the name
60 on the A-record for the host is still done for TA (but not for EE mode).
62 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
63 included in delivery lines for non-proxied connections, when compiled with
64 SUPPORT_SOCKS and running with proxy logging enabled.
70 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
71 When redis_servers is set to a list of > 1 element, and the Redis servers
72 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
73 case of MOVED into a DEFER case instead, thus moving the query onto the
74 next server in the list. For a cluster of N elements, all N servers must
75 be defined in redis_servers.
77 GF/02 Catch and remove uninitialized value warning in exiqsumm
78 Check for existence of @ARGV before looking at $ARGV[0]
80 JH/01 Replace the store_release() internal interface with store_newblock(),
81 which internalises the check required to safely use the old one, plus
82 the allocate and data copy operations duplicated in both (!) of the
85 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
86 modifier. This matches the restriction on the commandline.
88 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
89 Previously only the last row was returned.
91 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
92 we assumed that tags in the header were well-formed, and parsed the
93 element content after inspecting only the first char of the tag.
94 Assumptions at that stage could crash the receive process on malformed
97 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
98 While running the DKIM ACL we operate on the Permanent memory pool so that
99 variables created with "set" persist to the DATA ACL. Also (at any time)
100 DNS lookups that fail create cache records using the Permanent pool. But
101 expansions release any allocations made on the current pool - so a dnsdb
102 lookup expansion done in the DKIM ACL releases the memory used for the
103 DNS negative-cache, and bad things result. Solution is to switch to the
104 Main pool for expansions.
105 While we're in that code, add checks on the DNS cache during store_reset,
106 active in the testsuite.
107 Problem spotted, and debugging aided, by Wolfgang Breyha.
109 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
110 When none of the hosts presented to a transport match an already-open
111 connection, close it and proceed with the list. Previously we would
112 queue the message. Spotted by Lena with Yahoo, probably involving
115 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
116 Previously a spurious "250 OK id=" response was appended to the proper
119 JH/08 The "support for" informational output now, which built with Content
120 Scanning support, has a line for the malware scanner interfaces compiled
121 in. Interface can be individually included or not at build time.
123 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
124 by the template makefile "src/EDITME". The "STREAM" support for an older
125 ClamAV interface method is removed.
127 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
128 rows affected is given instead).
130 JH/11 The runtime Berkeley DB library version is now additionally output by
131 "exim -d -bV". Previously only the compile-time version was shown.
133 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
134 SMTP connection. Previously, when one had more receipients than the
135 first, an abortive onward connection was made. Move to full support for
136 multiple onward connections in sequence, handling cutthrough connection
137 for all multi-message initiating connections.
139 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
140 routers. Previously, a multi-recipient message would fail to match the
141 onward-connection opened for the first recipient, and cause its closure.
143 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
144 a timeout on read on a GnuTLS initiating connection, resulting in the
145 initiating connection being dropped. This mattered most when the callout
146 was marked defer_ok. Fix to keep the two timeout-detection methods
149 JH/15 Relax results from ACL control request to enable cutthrough, in
150 unsupported situations, from error to silently (except under debug)
151 ignoring. This covers use with PRDR, frozen messages, queue-only and
154 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
156 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
157 metadata, resulting in a crash in free().
159 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
160 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
161 Broken also in d185889f4, with init system revamp.
163 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
164 Previously we abruptly closed the connection after reading a malware-
165 found indication; now we go on to read the "scan ok" response line,
168 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
169 ACL. Previously, a crash would result.
171 JH/19 Speed up macro lookups during configuration file read, by skipping non-
172 macro text after a replacement (previously it was only once per line) and
173 by skipping builtin macros when searching for an uppercase lead character.
175 JH/20 DANE support moved from Experimental to mainline. The Makefile control
176 for the build is renamed.
178 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
179 was allocated for every new TLS startup, meaning one per message. Fix
180 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
182 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
183 reported the original. Fix to report (as far as possible) the ACL
184 result replacing the original.
186 JH/23 Fix memory leak during multi-message connections using STARTTLS under
187 OpenSSL. Certificate information is loaded for every new TLS startup,
188 and the resources needed to be freed.
190 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
192 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
195 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
196 DATA response info to the (existing) per-recipient response info for
197 the "C=" log element. It can have useful tracking info from the
198 destination system. Patch from Simon Arlott.
200 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
201 length value. Previously this would segfault.
203 HS/02 Support Avast multiline protoocol, this allows passing flags to
204 newer versions of the scanner.
206 JH/28 Ensure that variables possibly set during message acceptance are marked
207 dead before release of memory in the daemon loop. This stops complaints
208 about them when the debug_store option is enabled. Discovered specifically
209 for sender_rate_period, but applies to a whole set of variables.
210 Do the same for the queue-runner and queue-list loops, for variables set
211 from spool message files. Do the same for the SMTP per-message loop, for
212 certain variables indirectly set in ACL operations.
214 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
215 as a multi-recipient message from a mailinglist manager). The coding had
216 an arbitrary cutoff number of characters while checking for more input;
217 enforced by writing a NUL into the buffer. This corrupted long / fast
218 input. The problem was exposed more widely when more pipelineing of SMTP
219 responses was introduced, and one Exim system was feeding another.
220 The symptom is log complaints of SMTP syntax error (NUL chars) on the
221 receiving system, and refused recipients seen by the sending system
222 (propating to people being dropped from mailing lists).
223 Discovered and pinpointed by David Carter.
225 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
226 replaced by the ${authresults } expansion.
228 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
230 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
231 allows proper process termination in container environments.
233 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
234 Previously the "final dot" had a newline after it; ensure it is CR,LF.
236 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
237 and "err_perm", deprecated since 4.83 when the RFC-defined words
238 "temperror" and "permerror" were introduced.
240 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
241 transport-filters or DKIM-signing. The restriction was lost in the
242 consolidation of verify-callout and delivery SMTP handling.
243 Extend the restriction to also cover ARC-signing.
245 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
246 in defer=pass mode supply a 450 to the initiator. Previously the message
249 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
250 tls_require_ciphers is used as before.
252 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
253 "pass_unscanned". Only tmpfails from the scanner are written to
254 the paniclog, as they may require admin intervention (permission
255 denied, license issues). Other scanner errors (like decompression
256 bombs) do not cause a paniclog entry.
258 JH/36 Fix reinitialisation of DKIM logging variable between messages.
259 Previously it was possible to log spurious information in receive log
262 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
263 triggered odd behaviour from Outlook Express clients.
265 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
268 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
269 since the IETF WG has not yet settled on that versus the original
270 "bare" representation.
272 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
273 Previously the millisecond value corrupted the output.
274 Fix also for syslog_pid=no and log_selector +pid, for which the pid
275 corrupted the output.
281 JH/01 Rework error string handling in TLS interface so that the caller in
282 more cases is responsible for logging. This permits library-sourced
283 string to be attached to addresses during delivery, and collapses
284 pairs of long lines into single ones.
286 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
287 during configuration. Wildcards are allowed and expanded.
289 JH/02 Rework error string handling in DKIM to pass more info back to callers.
290 This permits better logging.
292 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
293 do not close it down and have the child transport start it up again on
294 the passed-on TCP connection. Instead, proxy the child (and any
295 subsequent ones) for TLS via a unix-domain socket channel. Logging is
296 affected: the continued delivery log lines do not have any DNSSEC, TLS
297 Certificate or OCSP information. TLS cipher information is still logged.
299 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
300 identical IP addresses on different listening ports. Will also affect
303 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
304 add noisy ifdef guards to special-case this sillyness.
305 Patch from Bernd Kuhls.
307 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
308 than 255 are no longer allowed.
310 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
311 Disable the session-cache too, which might reduce our load. Since we
312 currrectly use a new context for every connection, both as server and
313 client, there is no benefit for these.
314 GnuTLS appears to not support tickets server-side by default (we don't
315 call gnutls_session_ticket_enable_server()) but client side is enabled
316 by default on recent versions (3.1.3 +) unless the PFS priority string
319 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
320 <https://reproducible-builds.org/specs/source-date-epoch/>.
322 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
323 the check for any unsuccessful recipients did not notice the limit, and
324 erroneously found still-pending ones.
326 JH/08 Pipeline CHUNKING command and data together, on kernels that support
327 MSG_MORE. Only in-clear (not on TLS connections).
329 JH/09 Avoid using a temporary file during transport using dkim. Unless a
330 transport-filter is involved we can buffer the headers in memory for
331 creating the signature, and read the spool data file once for the
332 signature and again for transmission.
334 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
335 4.77 as the kernel support then wasn't solid, having issues in 64bit
336 mode. Now, it's been long enough. Add support for FreeBSD also.
338 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
339 case where the routing stage had gathered several addresses to send to
340 a host before calling the transport for the first, we previously failed
341 to close down TLS in the old transport process before passing the TCP
342 connection to the new process. The new one sent a STARTTLS command
343 which naturally failed, giving a failed delivery and bloating the retry
344 database. Investigation and fix prototype from Wolfgang Breyha.
346 JH/12 Fix check on SMTP command input synchronisation. Previously there were
347 false-negatives in the check that the sender had not preempted a response
348 or prompt from Exim (running as a server), due to that code's lack of
349 awareness of the SMTP input buffering.
351 PP/04 Add commandline_checks_require_admin option.
352 Exim drops privileges sanely, various checks such as -be aren't a
353 security problem, as long as you trust local users with access to their
354 own account. When invoked by services which pass untrusted data to
355 Exim, this might be an issue. Set this option in main configuration
356 AND make fixes to the calling application, such as using `--` to stop
359 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
360 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
361 responses to those, into a single TLS record each way (this usually means
362 a single packet). As a side issue, smtp_enforce_sync now works on TLS
365 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
366 affects you only if you're dancing at the edge of the param size limits.
367 If you are, and this message makes sense to you, then: raise the
368 configured limit or use OpenSSL 1.1. Nothing we can do for older
371 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
372 cmdline element to get the documented default one. Previously it was
375 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
376 are unsuable for non-message-specific reasons - eg. connection timeout,
377 banner-time rejection.
379 JH/16 Fix logging of delivery remote port, when specified by router, under
382 PP/06 Repair manualroute's ability to take options in any order, even if one
383 is the name of a transport.
386 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
388 JH/17 Change the list-building routines interface to use the expanding-string
389 triplet model, for better allocation and copying behaviour.
391 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
392 Previously it was constructed the first time a possibly-matching string
393 was met in the configuration file input during startup; now it is done
396 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
397 compatible one, to avoid the (poorly documented) possibility of a config
398 file in the working directory redirecting the DB files, possibly correpting
399 some existing file. CVE-2017-10140 assigned for BDB.
401 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
402 cache-hot. Previously, although the result was properly cached, the
403 initial verify call returned a defer.
405 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
406 the main verify for receipient in uncached-mode.
408 JH/22 Retire historical build files to an "unsupported" subdir. These are
409 defined as "ones for which we have no current evidence of testing".
411 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
412 if present. Previously it was ignored.
414 JH/24 Start using specified-initialisers in C structure init coding. This is
415 a C99 feature (it's 2017, so now considered safe).
417 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
418 if was a fixed-sized field and bitmask ops via macros; it is now more
421 PP/07 GitHub PR 56: Apply MariaDB build fix.
422 Patch provided by Jaroslav Škarvada.
424 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
425 during Coverity cleanups [4.87 JH/47]
426 Diagnosis and fix provided by Michael Fischer v. Mollard.
428 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
429 the right size to place the terminating semicolon on its own folded
430 line, the header hash was calculated to an incorrect value thanks to
431 the (relaxed) space the fold became.
433 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
434 and confused the parent.
436 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
437 which could crash as a result. This could lead to undeliverable messages.
439 JH/28 Logging: "next input sent too soon" now shows where input was truncated
442 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
443 matters on fast-turnover and PID-randomising systems, which were getting
444 out-of-order delivery.
446 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
447 a possibly-overlapping copy. The symptom was that "Remote host closed
448 connection in response to HELO" was logged instead of the actual 4xx
451 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
452 Previously only that bufferd was discarded, resulting in SYMTP command
455 JH/32 DKIM: when a message has multiple signatures matching an identity given
456 in dkim_verify_signers, run the dkim acl once for each. Previously only
457 one run was done. Bug 2189.
459 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
460 "panic the current process" to "deliberately defer". The panic log is
461 still written with the problem list name; the mail and reject logs now
462 get a temp-reject line for the message that was being handled, saying
463 something like "domains check lookup or other defer". The SMTP 451
464 message is still "Temporary local problem".
466 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
467 A crafted sequence of BDAT commands could result in in-use memory beeing
468 freed. CVE-2017-16943.
470 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
471 from SMTP input. Previously it was always done; now only done for DATA
472 and not BDAT commands. CVE-2017-16944.
474 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
475 to the message (such as an overlong header line). Previously this was
476 not done and we did not exit BDAT mode. Followon from the previous item
477 though a different problem.
483 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
484 than -2003 did; needs libidn2 in addition to libidn.
486 JH/02 The path option on a pipe transport is now expanded before use.
488 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
489 Patch provided by "Björn", documentation fix added too.
491 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
492 missing a wire-to-host endian conversion.
494 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
495 close after a BDAT command line could be taken as a following command,
496 giving a synch failure. Fix by only checking for synch immediately
497 before acknowledging the chunk.
499 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
500 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
501 macro. Patches provided by Josh Soref.
503 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
504 Previously we did not; the RFC seems ambiguous and VRFY is not listed
505 by IANA as a service extension. However, John Klensin suggests that we
508 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
509 the dkim code may be unix-mode line endings rather than smtp wire-format
510 CRLF, so prepend a CR to any bare LF.
512 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
513 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
515 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
516 the first were themselves being wrongly included in the feed into dkim
517 processing; with most chunk sizes in use this resulted in an incorrect
518 body hash calculated value.
520 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
521 DKIM signature block, for verification. Although advised against by
522 standards it is specifically not ruled illegal.
524 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
526 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
527 missing a body hash (the bh= tag).
529 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
530 It seems that HAProxy sends the Proxy Protocol information in clear and
531 only then does a TLS startup, so do the same.
533 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
534 TCP connections (such as for Spamd) unless the daemon successfully set
535 Fast Open mode on its listening sockets. This fixes breakage seen on
536 too-old kernels or those not configured for Fast Open, at the cost of
537 requiring both directions being enabled for TFO, and TFO never being used
538 by non-daemon-related Exim processes.
540 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
541 endings, at least on the first header line. Try to canonify any that get
542 past that check, despite the cost.
544 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
545 now limited to an arbitrary five deep, while parsing addresses with the
546 strip_excess_angle_brackets option enabled.
548 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
549 instead leave the unprompted TLS handshake in socket buffer for the
550 TLS library to consume.
552 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
554 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
556 JH/16 Drop variables when they go out of scope. Memory management drops a whole
557 region in one operation, for speed, and this leaves assigned pointers
558 dangling. Add checks run only under the testsuite which checks all
559 variables at a store-reset and panics on a dangling pointer; add code
560 explicitly nulling out all the variables discovered. Fixes one known
561 bug: a transport crash, where a dangling pointer for $sending_ip_address
562 originally assigned in a verify callout, is re-used.
564 PP/06 Drop '.' from @INC in various Perl scripts.
566 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
568 PP/08 Reduce a number of compilation warnings under clang; building with
569 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
570 should be warning-free.
572 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
574 HS/01 Fix portability problems introduced by PP/08 for platforms where
575 realloc(NULL) is not equivalent to malloc() [SunOS et al].
577 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
578 chunk. This allows us to accept broken chunked messages. We need a more
579 general solution here.
581 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
582 already-broken messages in the queue.
584 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
586 JH/19 Fix reference counting bug in routing-generated-address tracking.
592 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
593 supports it and a size is available (ie. the sending peer gave us one).
595 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
596 years of being deprecated). The replacements are the ACLs
597 acl_smtp_mime and acl_not_smtp_mime.
599 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
600 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
601 or even an in-clear connection were permitted. Now, if the host lookup
602 was dnssec and dane was requested then the host is only used if the
603 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
604 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
605 if one fails this test.
606 This means that a poorly-configured remote DNS will make it incommunicado;
607 but it protects against a DNS-interception attack on it.
609 JH/04 Bug 1810: make continued-use of an open smtp transport connection
610 non-noisy when a race steals the message being considered.
612 JH/05 If main configuration option tls_certificate is unset, generate a
613 self-signed certificate for inbound TLS connections.
615 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
616 in rewrites and routers.
618 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
619 and logged a warning sing 4.83; now they are a configuration file error.
621 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
622 (lacking @domain). Apply the same qualification processing as RCPT.
624 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
626 JH/10 Support ${sha256:} applied to a string (as well as the previous
629 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
630 a cutthrough deliver is pending, as we always want to make a connection.
631 This also avoids re-routing the message when later placing the cutthrough
632 connection after a verify cache hit.
633 Do not update it with the verify result either.
635 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
636 when routing results in more than one destination address.
638 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
639 signing (which inhibits the cutthrough capability). Previously only
640 the presence of an option was tested; now an expansion evaluating as
641 empty is permissible (obviously it should depend only on data available
642 when the cutthrough connection is made).
644 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
645 the relevant preceding SMTP command did not note the pipelining mode.
647 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
648 Previously they were not counted.
650 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
651 as one having no matching records. Previously we deferred the message
652 that needed the lookup.
654 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
655 distinguished as "(=".
657 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
658 for missing MX records. Previously it only worked for missing A records.
660 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
662 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
663 after the data-go-ahead and data-ack. Patch from Jason Betts.
665 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
666 even for a "none" policy. Patch from Tony Meyer.
668 JH/22 Fix continued use of a connection for further deliveries. If a port was
669 specified by a router, it must also match for the delivery to be
672 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
673 When one of the recipients of a message was unsuitable for the connection
674 (has no matching addresses), we lost track of needing to mark it
675 deferred. As a result mail would be lost.
677 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
679 JH/25 Decoding ACL controls is now done using a binary search; the source code
680 takes up less space and should be simpler to maintain. Merge the ACL
681 condition decode tables also, with similar effect.
683 JH/26 Fix problem with one_time used on a redirect router which returned the
684 parent address unchanged. A retry would see the parent address marked as
685 delivered, so not attempt the (identical) child. As a result mail would
688 JH/27 Fix a possible security hole, wherein a process operating with the Exim
689 UID can gain a root shell. Credit to http://www.halfdog.net/ for
690 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
693 JH/28 Enable {spool,log} filesystem space and inode checks as default.
694 Main config options check_{log,spool}_{inodes,space} are now
695 100 inodes, 10MB unless set otherwise in the configuration.
697 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
698 Previously it only applied to the main-section connection policy
701 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
703 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
704 by me. Added RFC7919 DH primes as an alternative.
706 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
707 are not in the system include path.
709 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
710 GnuTLS, when a session startup failed (eg because the client disconnected)
711 Exim did stdio operations after fclose. This was exposed by a recent
712 change which nulled out the file handle after the fclose.
714 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
715 signed directly by the cert-signing cert, rather than an intermediate
716 OCSP-signing cert. This is the model used by LetsEncrypt.
718 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
720 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
721 an incoming connection.
723 HS/02 Bug 1802: Do not half-close the connection after sending a request
726 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
727 fallback to "prime256v1".
729 JH/34 SECURITY: Use proper copy of DATA command in error message.
730 Could leak key material. Remotely exploitable. CVE-2016-9963.
736 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
737 and 3.4.4 - once the server is enabled to respond to an OCSP request
738 it does even when not requested, resulting in a stapling non-aware
739 client dropping the TLS connection.
741 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
742 support variable-length bit vectors. No functional change.
744 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
745 The I= interface field on outgoing lines is now after the H= remote
746 host field, same as incoming lines. There is a separate
747 outgoing_interface log selector which allows you to disable the
750 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
751 If not running log_selector +smtp_connection the mainlog would be held
752 open indefinitely after a "too many connections" event, including to a
753 deleted file after a log rotate. Leave the per net connection logging
754 leaving it open for efficiency as that will be quickly detected by the
755 check on the next write.
757 HS/01 Bug 1671: Fix post transport crash.
758 Processing the wait-<transport> messages could crash the delivery
759 process if the message IDs didn't exist for some reason. When
760 using 'split_spool_directory=yes' the construction of the spool
761 file name failed already, exposing the same netto behaviour.
763 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
764 mime_regex ACL conditions.
766 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
767 to DSN fail messages (bounces): remote IP, remote greeting, remote response
768 to HELO, local diagnostic string.
770 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
771 log line to debug. Even when configured with a tls authenticator many
772 client connections are expected to not authenticate in this way, so
773 an authenticate fail is not an error.
775 HS/02 Add the Exim version string to the process info. This way exiwhat
776 gives some more detail about the running daemon.
778 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
779 matter for fast-change records such as DNSBLs.
781 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
782 retry record, even if constant. There may be multiple transports with
783 different interface settings and the retry behaviour needs to be kept
786 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
788 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
790 JH/10 Bug 840: fix log_defer_output option of pipe transport
792 JH/11 Bug 830: use same host for all RCPTS of a message, even under
793 hosts_randomize. This matters a lot when combined with mua_wrapper.
795 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
796 ${quote_pgsql:<string>} operator.
798 JH/13 Bug 1708: avoid misaligned access in cached lookup.
800 JH/14 Change header file name for freeradius-client. Relevant if compiling
801 with Radius support; from the Gentoo tree and checked under Fedora.
803 JH/15 Bug 1712: Introduce $prdr_requested flag variable
805 JH/16 Bug 1714: Permit an empty string as expansion result for transport
806 option transport_filter, meaning no filtering.
808 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
810 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
811 defaults to "*" (all hosts). The variable is now available when not built
812 with TLS, default unset, mainly to enable keeping the testsuite sane.
813 If a server certificate is not supplied (via tls_certificate) an error is
814 logged, and clients will find TLS connections fail on startup. Presumably
815 they will retry in-clear.
816 Packagers of Exim are strongly encouraged to create a server certificate
817 at installation time.
819 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
820 with the $config_file variable.
822 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
823 in transport context, after the attempt, and per-recipient. The latter type
824 is per host attempted. The event data is the error message, and the errno
825 information encodes the lookup type (A vs. MX) used for the (first) host,
826 and the trailing two digits of the smtp 4xx response.
828 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
829 to write to mainlog (or rejectlog, paniclog) in the window between file
830 creation and permissions/ownership being changed. Particularly affects
831 installations where exicyclog is run as root, rather than exim user;
832 result is that the running daemon panics and dies.
834 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
836 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
837 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
838 "pri" and "weight". Note that the previous implicit priority given by the
839 list order is no longer honoured.
841 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
844 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
845 by defining SUPPORT_SOCKS.
847 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
848 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
849 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
850 variables are renamed to proxy_{local,external}_{address,port}.
852 JH/27 Move Internationalisation support from Experimental to mainline, enabled
853 for a build by defining SUPPORT_I18N
855 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
856 of the query string, and make ${quote_redis:} do that quoting.
858 JH/29 Move Events support from Experimental to mainline, enabled by default
859 and removable for a build by defining DISABLE_EVENT.
861 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
863 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
864 cached by the daemon.
866 JH/32 Move Redis support from Experimental to mainline, enabled for a build
867 by defining LOOKUP_REDIS. The libhiredis library is required.
869 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
870 keys are given for lookup.
872 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
873 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
874 only supported when built with TLS support. The PolarSSL SHA routines
875 are still used when the TLS library is too old for convenient support.
877 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
878 openssl_options), for security. OpenSSL forces this from version 1.1.0
879 server-side so match that on older versions.
881 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
882 allocation for $value could be released as the expansion processing
883 concluded, but leaving the global pointer active for it.
885 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
886 and to use the domains and local_parts ACL conditions.
888 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
889 incorrectly not doubled on cutthrough transmission, hence seen as a
890 body-termination at the receiving system - resulting in truncated mails.
891 Commonly the sender saw a TCP-level error, and retransmitted the message
892 via the normal store-and-forward channel. This could result in duplicates
893 received - but deduplicating mailstores were liable to retain only the
894 initial truncated version.
896 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
898 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
900 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
901 we're in there, support oversigning also; bug 1309.
903 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
905 HS/04 Add support for keep_environment and add_environment options.
907 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
908 either intentional arithmetic overflow during PRNG, or testing config-
911 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
912 delivery resulted in actual delivery. Cancel cutthrough before DATA
915 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
916 rejects a recipient: pass the reject to the originator.
918 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
919 Many were false-positives and ignorable, but it's worth fixing the
922 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
923 for the new environment-manipulation done at startup. Move the routines
924 from being local to tls.c to being global via the os.c file.
926 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
927 an extract embedded as result-arg for a map, the first arg for extract
928 is unavailable so we cannot tell if this is a numbered or keyed
929 extraction. Accept either.
935 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
938 JH/02 The smtp transport option "multi_domain" is now expanded.
940 JH/03 The smtp transport now requests PRDR by default, if the server offers
943 JH/04 Certificate name checking on server certificates, when exim is a client,
944 is now done by default. The transport option tls_verify_cert_hostnames
945 can be used to disable this per-host. The build option
946 EXPERIMENTAL_CERTNAMES is withdrawn.
948 JH/05 The value of the tls_verify_certificates smtp transport and main options
949 default to the word "system" to access the system default CA bundle.
950 For GnuTLS, only version 3.0.20 or later.
952 JH/06 Verification of the server certificate for a TLS connection is now tried
953 (but not required) by default. The verification status is now logged by
954 default, for both outbound TLS and client-certificate supplying inbound
957 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
960 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
961 Status Notification (bounce) messages are now MIME format per RFC 3464.
962 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
963 under the control of the dsn_advertise_hosts option, and routers may
964 have a dsn_lasthop option.
966 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
967 default, modifiable by a malware= option. The list separator for
968 the options can now be changed in the usual way. Bug 68.
970 JH/10 The smtp_receive_timeout main option is now expanded before use.
972 JH/11 The incoming_interface log option now also enables logging of the
973 local interface on delivery outgoing connections.
975 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
976 if the interface and destination host and port all match.
978 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
981 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
982 Patch from Andrew Lewis.
984 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
985 now supports optional time-restrictions, weighting, and priority
986 modifiers per server. Patch originally by <rommer@active.by>.
988 JH/16 The spamd_address main option now supports a mixed list of local
989 and remote servers. Remote servers can be IPv6 addresses, and
990 specify a port-range.
992 JH/17 Bug 68: The spamd_address main option now supports an optional
993 timeout value per server.
995 JH/18 Bug 1581: Router and transport options headers_add/remove can
996 now have the list separator specified.
998 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1001 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1004 JH/21 Support for the A6 type of dns record is withdrawn.
1006 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1007 rather than the verbs used.
1009 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1010 from 255 to 1024 chars.
1012 JH/24 Verification callouts now attempt to use TLS by default.
1014 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1015 are generic router options now. The defaults didn't change.
1017 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1018 Original patch from Alexander Shikoff, worked over by JH.
1020 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1021 files if ZCAT_COMMAND is not executable.
1023 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1025 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1027 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1028 Normally benign, it bites when the pair was led to by a CNAME;
1029 modern usage is to not canonicalize the domain to a CNAME target
1030 (and we were inconsistent anyway for A-only vs AAAA+A).
1032 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1034 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1035 when evaluating $sender_host_dnssec.
1037 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1038 $sender_helo_dnssec variable.
1040 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1042 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1044 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1046 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1047 documented as working, but never had. Support all but $spam_report.
1049 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1050 added for tls authenticator.
1052 HS/03 Add perl_taintmode main config option
1058 TL/01 When running the test suite, the README says that variables such as
1059 no_msglog_check are global and can be placed anywhere in a specific
1060 test's script, however it was observed that placement needed to be near
1061 the beginning for it to behave that way. Changed the runtest perl
1062 script to read through the entire script once to detect and set these
1063 variables, reset to the beginning of the script, and then run through
1064 the script parsing/test process like normal.
1066 TL/02 The BSD's have an arc4random API. One of the functions to induce
1067 adding randomness was arc4random_stir(), but it has been removed in
1068 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1069 function when detected.
1071 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1072 cause callback expansion.
1074 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1075 syntax errors in an expansion can be treated as a string instead of
1076 logging or causing an error, due to the internal use of bool_lax
1077 instead of bool when processing it.
1079 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1080 server certificates when making smtp deliveries.
1082 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1084 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1086 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1088 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1089 Merged patch from Sebastian Wiedenroth.
1091 JH/05 Fix results-pipe from transport process. Several recipients, combined
1092 with certificate use, exposed issues where response data items split
1093 over buffer boundaries were not parsed properly. This eventually
1094 resulted in duplicates being sent. This issue only became common enough
1095 to notice due to the introduction of connection certificate information,
1096 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1098 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1099 size buffer was used, resulting in syntax errors when an expansion
1102 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1103 version 3.3.6 or later.
1105 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1106 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1107 both become "event_action", the variables become $event_name, $event_data
1108 and $event_defer_errno. There is a new variable $verify_mode, usable in
1109 routers, transports and related events. The tls:cert event is now also
1110 raised for inbound connections, if the main configuration event_action
1113 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1114 early OCSP support, but no stapling (appears to be less than 1.0.0).
1116 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1117 server certificate names available under the smtp transport option
1118 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1121 JH/10 Time-related extraction expansions from certificates now use the main
1122 option "timezone" setting for output formatting, and are consistent
1123 between OpenSSL and GnuTLS compilations. Bug 1541.
1125 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1126 encoded parameter in the incoming message. Bug 1558.
1128 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1129 include certificate info, eximon was claiming there were spoolfile
1132 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1134 JH/14 Log delivery-related information more consistently, using the sequence
1135 "H=<name> [<ip>]" wherever possible.
1137 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1138 are problematic for Debian distribution, omit them from the release
1141 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1143 JH/16 Fix string representation of time values on 64bit time_t architectures.
1146 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1147 output list separator was used.
1152 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1153 checkers that were complaining about end of non-void function with no
1156 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1157 This was a regression introduced in 4.83 by another bugfix.
1159 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1161 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1162 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1168 TF/01 Correctly close the server side of TLS when forking for delivery.
1170 When a message was received over SMTP with TLS, Exim failed to clear up
1171 the incoming connection properly after forking off the child process to
1172 deliver the message. In some situations the subsequent outgoing
1173 delivery connection happened to have the same fd number as the incoming
1174 connection previously had. Exim would try to use TLS and fail, logging
1175 a "Bad file descriptor" error.
1177 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1178 utilities have not been installed.
1180 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1181 temporary space as the ACL may create new global variables.
1183 TL/01 LDAP support uses per connection or global context settings, depending
1184 upon the detected version of the libraries at build time.
1186 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1187 to extract and use the src ip:port in logging and expansions as if it
1188 were a direct connection from the outside internet. PPv2 support was
1189 updated based on HAProxy spec change in May 2014.
1191 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1193 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1194 Properly escape header and check for NULL return.
1196 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1199 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1201 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1202 characters in header names, implemented as a verify condition.
1203 Contributed by Michael Fischer v. Mollard.
1205 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1206 results permerror and temperror. Previous values are deprecated but
1207 still accepted. In a future release, err_perm and err_temp will be
1208 completely removed, which will be a backward incompatibility if the
1209 ACL tests for either of these two old results. Patch contributed by
1210 user bes-internal on the mailing list.
1212 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1214 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1215 selectors, in both main and reject logs.
1217 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1220 JH/07 Add malware type "sock" for talking to simple daemon.
1222 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1224 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1225 routers/transports under cutthrough routing.
1227 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1228 numbers. Touch up "bool" conditional to keep the same definition.
1230 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1232 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1234 JH/12 Expand items in router/transport headers_add or headers_remove lists
1235 individually rather than the list as a whole. Bug 1452.
1237 Required for reasonable handling of multiple headers_ options when
1238 they may be empty; requires that headers_remove items with embedded
1239 colons must have them doubled (or the list-separator changed).
1241 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1242 view the policy declared in the DMARC record. Currently, $dmarc_status
1243 is a combined value of both the record presence and the result of the
1246 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1248 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1249 dnslookup router and the smtp transport (applying to the forward
1252 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1253 of ldap servers used for a specific lookup. Patch provided by Heiko
1256 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1257 New variable $lookup_dnssec_authenticated for observability.
1259 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1260 Patch submitted by Lars Timman.
1262 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1264 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1265 Requires trusted mode and valid format message id, aborts otherwise.
1266 Patch contributed by Heiko Schlichting.
1268 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1269 certextract with support for various fields. Bug 1358.
1271 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1272 is requested by default, modifiable by smtp transport option
1275 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1276 operate on certificate variables to give certificate fingerprints
1277 Also new ${sha256:cert_variable}.
1279 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1281 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1282 Christian Aistleitner.
1284 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1286 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1287 file. Patch from Wolfgang Breyha.
1289 JH/25 Expand the coverage of the delivery $host and $host_address to
1290 client authenticators run in verify callout. Bug 1476.
1292 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1293 align with daemon_smtp_ports. Bug 72.
1295 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1296 support and error reporting did not work properly.
1298 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1299 and is readable. Patch from Andrew Colin Kissa.
1301 TL/14 Enhance documentation of ${run expansion and how it parses the
1302 commandline after expansion, particularly in the case when an
1303 unquoted variable expansion results in an empty value.
1305 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1307 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1308 by renaming away from it. Addresses GH issue 15, reported by
1311 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1313 TL/15 SECURITY: prevent double expansion in math comparison functions
1314 (can expand unsanitized data). Not remotely exploitable.
1321 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1323 PP/02 Make -n do something, by making it not do something.
1324 When combined with -bP, the name of an option is not output.
1326 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1329 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1330 $sender_host_name and config options to manage this, and basic check
1333 PP/05 DSCP support for outbound connections and control modifier for inbound.
1335 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1336 (Only plugin which currently uses this is kerberos4, which nobody should
1337 be using, but we should make it available and other future plugins might
1338 conceivably use it, even though it would break NAT; stuff *should* be
1339 using channel bindings instead).
1341 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1342 name; added for Sendmail compatibility; requires admin caller.
1343 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1344 just ignore it); requires trusted caller.
1345 Also parse but ignore: -Ac -Am -X<logfile>
1348 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1350 TL/02 Add +smtp_confirmation as a default logging option.
1352 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1353 Patch by Magnus Holmgren from 2007-02-20.
1355 TL/04 Bugzilla 1281 - Spec typo.
1356 Bugzilla 1283 - Spec typo.
1357 Bugzilla 1290 - Spec grammar fixes.
1359 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1361 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1363 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1364 mailing list by by Dmitry Isaikin.
1366 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1368 JH/02 Support "G" suffix to numbers in ${if comparisons.
1370 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1372 NM/01 Bugzilla 1197 - Spec typo
1373 Bugzilla 1196 - Spec examples corrections
1375 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1377 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1378 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1381 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1382 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1384 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1385 "acl {{name}{arg}...}", and optional args on acl condition
1388 JH/05 Permit multiple router/transport headers_add/remove lines.
1390 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1392 JH/07 Avoid using a waiting database for a single-message-only transport.
1393 Performance patch from Paul Fisher. Bugzilla 1262.
1395 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1398 JH/09 Add $headers_added variable, with content from use of ACL modifier
1399 add_header (but not yet added to the message). Bugzilla 199.
1401 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1402 Pulled from Bugzilla 817 by Wolfgang Breyha.
1404 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1406 (nb: this is the same fix as in Exim 4.80.1)
1408 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1411 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1414 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1416 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1417 advertise SMTP AUTH mechanism to us, instead of a generic
1418 protocol violation error. Also, make Exim more robust to bad
1419 data from the Dovecot auth socket.
1421 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1423 When a queue runner is handling a message, Exim first routes the
1424 recipient addresses, during which it prunes them based on the retry
1425 hints database. After that it attempts to deliver the message to
1426 any remaining recipients. It then updates the hints database using
1429 So if a recipient address works intermittently, it can get repeatedly
1430 deferred at routing time. The retry hints record remains fresh so the
1431 address never reaches the final cutoff time.
1433 This is a fairly common occurrence when a user is bumping up against
1434 their storage quota. Exim had some logic in its local delivery code
1435 to deal with this. However it did not apply to per-recipient defers
1436 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1438 This change adds a proper retry rule check during routing so that the
1439 final cutoff time is checked against the message's age. We only do
1440 this check if there is an address retry record and there is not a
1441 domain retry record; this implies that previous attempts to handle
1442 the address had the retry_use_local_parts option turned on. We use
1443 this as an approximation for the destination being like a local
1444 delivery, as in LMTP.
1446 I suspect this new check makes the old local delivery cutoff check
1447 redundant, but I have not verified this so I left the code in place.
1449 TF/02 Correct gecos expansion when From: is a prefix of the username.
1451 Test 0254 submits a message to Exim with the header
1455 When I ran the test suite under the user fanf2, Exim expanded
1456 the header to contain my full name, whereas it should have added
1457 a Resent-Sender: header. It erroneously treats any prefix of the
1458 username as equal to the username.
1460 This change corrects that bug.
1462 GF/01 DCC debug and logging tidyup
1463 Error conditions log to paniclog rather than rejectlog.
1464 Debug lines prefixed by "DCC: " to remove any ambiguity.
1466 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1468 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1469 Bug spotted by Jeremy Harris; was flawed since initial commit.
1470 Would have resulted in OCSP responses post-SNI triggering an Exim
1471 NULL dereference and crash.
1473 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1475 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1476 Bug detection, analysis and fix by Samuel Thibault.
1477 Bugzilla 1331, Debian bug #698092.
1479 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1481 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1482 Server implementation by Todd Lyons, client by JH.
1483 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1484 config variable "prdr_enable" controls whether the server
1485 advertises the facility. If the client requests PRDR a new
1486 acl_data_smtp_prdr ACL is called once for each recipient, after
1487 the body content is received and before the acl_smtp_data ACL.
1488 The client is controlled by both of: a hosts_try_prdr option
1489 on the smtp transport, and the server advertisement.
1490 Default client logging of deliveries and rejections involving
1491 PRDR are flagged with the string "PRDR".
1493 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1494 fclose(). Diagnosis by Todd Lyons.
1496 PP/17 Update configure.default to handle IPv6 localhost better.
1497 Patch by Alain Williams (plus minor tweaks).
1500 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1501 This is now consistent with GnuTLS, and is now documented: the
1502 previous undocumented portable approach to treating the option as
1503 unset was to force an expansion failure. That still works, and
1504 an empty string is now equivalent.
1506 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1507 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1508 not performing validation itself.
1510 PP/20 Added force_command boolean option to pipe transport.
1511 Patch from Nick Koston, of cPanel Inc.
1513 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1516 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1518 PP/21 Fix eximon continuous updating with timestamped log-files.
1519 Broken in a format-string cleanup in 4.80, missed when I repaired the
1520 other false fix of the same issue.
1521 Report and fix from Heiko Schlichting.
1524 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1525 Report from Prashanth Katuri.
1527 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1528 It's SecureTransport, so affects any MacOS clients which use the
1529 system-integrated TLS libraries, including email clients.
1531 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1532 using a MIME ACL for non-SMTP local injection.
1533 Report and assistance in diagnosis by Warren Baker.
1535 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1537 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1539 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1540 last id that failed so it may be referenced in subsequent ACL's.
1542 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1545 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1546 ldap library initialization, allowing self-signed CA's to be
1547 used. Also properly sets require_cert option later in code by
1548 using NULL (global ldap config) instead of ldap handle (per
1549 session). Bug diagnosis and testing by alxgomz.
1551 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1552 the src/util/ subdirectory.
1554 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1555 renamed to Transport Post Delivery Action by Jeremy Harris, as
1558 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1559 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1560 redis_servers = needs to be configured which will be used by the redis
1561 lookup. Patch from Warren Baker, of The Packet Hub.
1563 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1565 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1566 hostname or reverse DNS when processing a host list. Used suggestions
1567 from multiple comments on this bug.
1569 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1571 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1572 Missed a few lines, added it to make the runtest require no keyboard
1575 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1576 contains upper case chars. Make router use caseful_local_part.
1578 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1579 support when GnuTLS has been built with p11-kit.
1585 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1587 This, or similar/improved, will also be change PP/11 of 4.82.
1593 PP/01 Handle short writes when writing local log-files.
1594 In practice, only affects FreeBSD (8 onwards).
1595 Bugzilla 1053, with thanks to Dmitry Isaikin.
1597 NM/01 Bugzilla 949 - Documentation tweak
1599 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1602 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1604 PP/02 Implemented gsasl authenticator.
1606 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1608 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1609 `pkg-config foo` for cflags/libs.
1611 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1612 with rest of GSASL and with heimdal_gssapi.
1614 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1615 `pkg-config foo` for cflags/libs for the TLS implementation.
1617 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1618 properties get this fed in as external SSF. A number of robustness
1619 and debugging improvements to the cyrus_sasl authenticator.
1621 PP/08 cyrus_sasl server now expands the server_realm option.
1623 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1624 Patch by Jeremy Harris.
1626 PP/10 Added dbmjz lookup type.
1628 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1630 PP/12 MAIL args handles TAB as well as SP, for better interop with
1631 non-compliant senders.
1632 Analysis and variant patch by Todd Lyons.
1634 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1635 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1636 Patch from Dirk Mueller <dmueller@suse.com>
1638 PP/13 tls_peerdn now print-escaped for spool files.
1639 Observed some $tls_peerdn in wild which contained \n, which resulted
1640 in spool file corruption.
1642 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1643 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1644 or write after TLS renegotiation, which otherwise led to messages
1647 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1648 as a tracking header (ie: a signed header comes before the signature).
1649 Patch from Wolfgang Breyha.
1651 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1652 comma-sep list; embedded commas doubled.
1654 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1656 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1658 Report and patch from Dmitry Banschikov.
1660 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1661 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1662 failures. We appear to now support TLS1.1+ with Exim.
1664 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1665 lets Exim select keys and certificates based upon TLS SNI from client.
1666 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1667 before an outbound SMTP session. New log_selector, +tls_sni.
1669 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1670 NULL dereference. Report and patch from Alun Jones.
1672 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1673 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1674 Not seeing resolver debug output on NetBSD, but suspect this is a
1675 resolver implementation change.
1677 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1678 Left warnings. Added "eximon gdb" invocation mode.
1680 PP/21 Defaulting "accept_8bitmime" to true, not false.
1682 PP/22 Added -bw for inetd wait mode support.
1684 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1685 locate the relevant includes and libraries. Made this the default.
1687 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1688 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1690 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1691 This may cause build issues on older platforms.
1693 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1694 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1695 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1696 Added SNI support via GnuTLS too.
1697 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1699 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1701 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1702 Applied second patch from Janne, implementing suggestion to default
1703 multiple-strings-in-record handling to match SPF spec.
1705 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1707 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1708 read-only, out of scope).
1709 Patch from Wolfgang Breyha, report from Stuart Northfield.
1711 PP/29 Fix three issues highlighted by clang analyser static analysis.
1712 Only crash-plausible issue would require the Cambridge-specific
1713 iplookup router and a misconfiguration.
1714 Report from Marcin Mirosław.
1716 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1718 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1719 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1720 As part of this, removing so much warning spew let me fix some minor
1721 real issues in debug logging.
1723 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1724 assignment on my part. Fixed.
1726 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1727 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1728 Janne Snabb (who went above and beyond: thank you).
1730 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1731 string otherwise requires a connection and a bunch more work and it's
1732 relatively easy to get wrong. Should also expose TLS library linkage
1735 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1736 64-bit ${eval} (JH/03).
1738 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1739 GNU libc to support some of the 64-bit stuff, should not lead to
1740 conflicts. Defined before os.h is pulled in, so if a given platform
1741 needs to override this, it can.
1743 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1744 protection layer was required, which is not implemented.
1745 Bugzilla 1254, patch from Wolfgang Breyha.
1747 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1748 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1749 tls_dhparam take prime identifiers. Also unbreak combination of
1750 OpenSSL+DH_params+TLSSNI.
1752 PP/39 Disable SSLv2 by default in OpenSSL support.
1758 PP/01 Solaris build fix for Oracle's LDAP libraries.
1759 Bugzilla 1109, patch from Stephen Usher.
1761 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1763 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1766 TF/02 Fix a couple more cases where we did not log the error message
1767 when unlink() failed. See also change 4.74-TF/03.
1769 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1770 lock up or crash if it happened to be inside a call to libc when it
1771 got a SIGUSR1 from exiwhat.
1773 The SIGUSR1 handler appends the current process status to the process
1774 log which is later printed by exiwhat. It used to use the general
1775 purpose logging code to do this, but several functions it calls are
1776 not safe for signals.
1778 The new output code in the SIGUSR1 handler is specific to the process
1779 log, and simple enough that it's easy to inspect for signal safety.
1780 Removing some special cases also simplifies the general logging code.
1781 Removing the spurious timestamps from the process log simplifies
1784 TF/04 Improved ratelimit ACL condition.
1786 The /noupdate option has been deprecated in favour of /readonly which
1787 has clearer semantics. The /leaky, /strict, and /readonly update modes
1788 are mutually exclusive. The update mode is no longer included in the
1789 database key; it just determines when the database is updated. (This
1790 means that when you upgrade Exim will forget old rate measurements.)
1792 Exim now checks that the per_* options are used with an update mode that
1793 makes sense for the current ACL. For example, when Exim is processing a
1794 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1795 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1796 must specify per_mail/readonly. If you omit the update mode it defaults to
1797 /leaky where that makes sense (as before) or /readonly where required.
1799 The /noupdate option is now undocumented but still supported for
1800 backwards compatibility. It is equivalent to /readonly except that in
1801 ACLs where /readonly is required you may specify /leaky/noupdate or
1802 /strict/noupdate which are treated the same as /readonly.
1804 A useful new feature is the /count= option. This is a generalization
1805 of the per_byte option, so that you can measure the throughput of other
1806 aggregate values. For example, the per_byte option is now equivalent
1807 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1809 The per_rcpt option has been generalized using the /count= mechanism
1810 (though it's more complicated than the per_byte equivalence). When it is
1811 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1812 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1813 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1814 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1815 Note that using per_rcpt with a non-readonly update mode in more than
1816 one ACL will cause the recipients to be double-counted. (The per_mail
1817 and per_byte options don't have this problem.)
1819 The handling of very low rates has changed slightly. If the computed rate
1820 is less than the event's count (usually one) then this event is the first
1821 after a long gap. In this case the rate is set to the same as this event's
1822 count, so that the first message of a spam run is counted properly.
1824 The major new feature is a mechanism for counting the rate of unique
1825 events. The new per_addr option counts the number of different
1826 recipients that someone has sent messages to in the last time period. It
1827 behaves like per_rcpt if all the recipient addresses are different, but
1828 duplicate recipient addresses do not increase the measured rate. Like
1829 the /count= option this is a general mechanism, so the per_addr option
1830 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1831 example, measure the rate that a client uses different sender addresses
1832 with the options per_mail/unique=$sender_address. There are further
1833 details in the main documentation.
1835 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1837 TF/06 Removed a few PCRE remnants.
1839 TF/07 Automatically extract Exim's version number from tags in the git
1840 repository when doing development or release builds.
1842 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1843 Bugzilla 879. Patch from Paul Fisher.
1845 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1846 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1849 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1851 PP/05 Variable $av_failed, true if the AV scanner deferred.
1852 Bugzilla 1078. Patch from John Horne.
1854 PP/06 Stop make process more reliably on build failure.
1855 Bugzilla 1087. Patch from Heiko Schlittermann.
1857 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1858 Bugzilla 1089. Patch from Heiko Schlittermann.
1860 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1861 Bugzilla 1131. Patch from Holger Weiß.
1863 PP/09 Handle IPv6 addresses with SPF.
1864 Bugzilla 860. Patch from Wolfgang Breyha.
1866 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1868 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1871 PP/11 match_* no longer expand right-hand-side by default.
1872 New compile-time build option, EXPAND_LISTMATCH_RHS.
1873 New expansion conditions, "inlist", "inlisti".
1875 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1877 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1879 PP/14 fix log_write() format string regression from TF/03.
1880 Bugzilla 1152. Patch from Dmitry Isaikin.
1886 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1888 PP/02 Harmonised TLS library version reporting; only show if debugging.
1889 Layout now matches that introduced for other libraries in 4.74 PP/03.
1891 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1893 PP/04 New "dns_use_edns0" global option.
1895 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1898 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1899 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1901 TK/01 Updated PolarSSL code to 0.14.2.
1902 Bugzilla 1097. Patch from Andreas Metzler.
1904 PP/07 Catch divide-by-zero in ${eval:...}.
1905 Fixes bugzilla 1102.
1907 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1910 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1911 format-string attack -- SECURITY: remote arbitrary code execution.
1913 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1914 time unintentionally subject to list matching rules, letting the header
1915 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1916 arbitrary string expansion). This allowed for information disclosure.
1918 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1919 INT_MIN/-1 -- value coerced to INT_MAX.
1925 NM/01 Workaround for PCRE version dependency in version reporting
1928 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1929 This fixes portability to compilers other than gcc, notably
1930 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1932 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1933 makefiles for portability to HP-UX and POSIX correctness.
1935 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1936 Also via indented variable definition in the Makefile.
1937 (Debugging by Oliver Heesakkers).
1939 PP/02 Restore caching of spamd results with expanded spamd_address.
1940 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1942 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1943 Improves build reliability. Fix from: Frank Elsner
1945 NM/02 Fix wide character breakage in the rfc2047 coding
1946 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1948 NM/03 Allow underscore in dnslist lookups
1949 Fixes bug 1026. Patch from Graeme Fowler
1951 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1952 Code patches from Adam Ciarcinski of NetBSD.
1954 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1957 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1958 is logged, to avoid truncation. Patch from John Horne.
1960 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1961 Patch from Jakob Hirsch.
1963 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1964 SQL string expansion failure details.
1965 Patch from Andrey Oktyabrski.
1967 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1968 Patch from Simon Arlott.
1970 PP/09 New lookups functionality failed to compile on old gcc which rejects
1971 extern declarations in function scope.
1972 Patch from Oliver Fleischmann
1974 PP/10 Use sig_atomic_t for flags set from signal handlers.
1975 Check getgroups() return and improve debugging.
1976 Fixed developed for diagnosis in bug 927 (which turned out to be
1979 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1980 Patch from Mark Zealey.
1982 PP/12 Bugzilla 1056: Improved spamd server selection.
1983 Patch from Mark Zealey.
1985 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1986 Based on patch from Heiko Schlittermann.
1988 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1989 Patch from Uwe Doering, sign-off by Michael Haardt.
1991 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1992 variable declaration deep within a block. Bug and patch from
1995 PP/15 lookups-Makefile IRIX compatibility coercion.
1997 PP/16 Make DISABLE_DKIM build knob functional.
1999 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2000 Patch by Simon Arlott
2002 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2003 variable argument macros. Our copy now differs from upstream.
2009 TF/01 Failure to get a lock on a hints database can have serious
2010 consequences so log it to the panic log.
2012 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2013 controlled using the smtp_confirmation log selector.
2015 TF/03 Include the error message when we fail to unlink a spool file.
2017 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2018 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2019 for maintaining out-of-tree patches for some time.
2021 PP/01 Bugzilla 139: Documentation and portability issues.
2022 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2023 Handle per-OS dynamic-module compilation flags.
2025 PP/02 Let /dev/null have normal permissions.
2026 The 4.73 fixes were a little too stringent and complained about the
2027 permissions on /dev/null. Exempt it from some checks.
2028 Reported by Andreas M. Kirchwitz.
2030 PP/03 Report version information for many libraries, including
2031 Exim version information for dynamically loaded libraries. Created
2032 version.h, now support a version extension string for distributors
2033 who patch heavily. Dynamic module ABI change.
2035 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2036 privilege escalation vulnerability whereby the Exim run-time user
2037 can cause root to append content of the attacker's choosing to
2040 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2043 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2044 If dropping privileges for untrusted macros, we disabled normal logging
2045 on the basis that it would fail; for the Exim run-time user, this is not
2046 the case, and it resulted in successful deliveries going unlogged.
2047 Fixed. Reported by Andreas Metzler.
2053 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2054 only prepend for the Resent-* case. Fixes regression introduced in
2055 Exim 4.70 by NM/22 for Bugzilla 607.
2057 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2058 increasing numbers of administrators be bitten by this.
2060 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2062 PP/03 Bugzilla 994: added openssl_options main configuration option.
2064 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2066 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2068 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2070 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2071 without a peer certificate, leading to a segfault because of an
2072 assumption that peers always have certificates. Be a little more
2073 paranoid. Problem reported by Martin Tscholak.
2075 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2076 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2077 NB: ClamAV planning to remove STREAM in "middle of 2010".
2078 CL also introduces -bmalware, various -d+acl logging additions and
2079 more caution in buffer sizes.
2081 PP/09 Implemented reverse_ip expansion operator.
2083 PP/10 Bugzilla 937: provide a "debug" ACL control.
2085 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2087 PP/12 Bugzilla 973: Implement --version.
2089 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2091 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2093 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2095 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2096 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2097 ignore trailing whitespace.
2099 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2101 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2104 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2105 Notification from Dr Andrew Aitchison.
2107 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2108 ExtendedDetectionInfo response format.
2109 Notification from John Horne.
2111 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2114 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2115 XSL and documented dependency on system catalogs, with examples of how
2118 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2121 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2122 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2123 configuration file which is writeable by the Exim user or group.
2125 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2126 of configuration files to cover files specified with the -C option if
2127 they are going to be used with root privileges, not just the default
2130 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2131 option (effectively making it always true).
2133 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2134 files to be used while preserving root privileges.
2136 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2137 that rogue child processes cannot use them.
2139 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2140 run-time user, instead of root.
2142 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2143 Exim run-time user without dropping privileges.
2145 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2146 result string, instead of calling string_vformat() twice with the same
2149 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2150 for other users. Others should always drop root privileges if they use
2151 -C on the command line, even for a whitelisted configure file.
2153 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2155 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2161 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2162 $data_path, and $header_path variables; fixed documentation bugs and
2165 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2166 exipick to access non-standard spools, including the "frozen" queue
2169 NM/01 Bugzilla 965: Support mysql stored procedures.
2170 Patch from Alain Williams
2172 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2174 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2175 Patch from Andreas Metzler
2177 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2178 Patch from Kirill Miazine
2180 NM/05 Bugzilla 671: Added umask to procmail example.
2182 JJ/03 installed exipick 20100323.0, fixing doc bug
2184 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2185 directory. Notification and patch from Dan Rosenberg.
2187 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2189 TK/02 Improve log output when DKIM signing operation fails.
2191 MH/01 Treat the transport option dkim_domain as a colon separated
2192 list, not as a single string, and sign the message with each element,
2193 omitting multiple occurences of the same signer.
2195 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2196 Bugzilla 985, 986. Patch by Simon Arlott
2198 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2199 Patch by Simon Arlott
2201 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2202 MBX locking. Notification from Dan Rosenberg.
2208 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2210 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2212 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2214 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2216 NM/04 Bugzilla 915: Flag broken perl installation during build.
2222 TK/01 Added patch by Johannes Berg that expands the main option
2223 "spamd_address" if it starts with a dollar sign.
2225 TK/02 Write list of recipients to X-Envelope-Sender header when building
2226 the mbox-format spool file for content scanning (suggested by Jakob
2229 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2230 (http://www.dcc-servers.net/) support via dccifd. Activated by
2231 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2233 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2234 by Mark Daniel Reidel <mr@df.eu>.
2236 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2237 When building exim an external PCRE library is now needed -
2238 PCRE is a system library on the majority of modern systems.
2239 See entry on PCRE_LIBS in EDITME file.
2241 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2242 conversation. Added nologin parameter to request.
2243 Patch contributed by Kirill Miazine.
2245 TF/01 Do not log submission mode rewrites if they do not change the address.
2247 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2249 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2250 log files in place. Contributed by Roberto Lima.
2252 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2254 TF/03 Bugzilla 615: When checking the local_parts router precondition
2255 after a local_part_suffix or local_part_prefix option, Exim now
2256 does not use the address's named list lookup cache, since this
2257 contains cached lookups for the whole local part.
2259 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2260 Robert Millan. Documentation is in experimental-spec.txt.
2262 TF/04 Bugzilla 668: Fix parallel build (make -j).
2264 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2266 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2267 Patch provided by Jan Srzednicki.
2269 TF/05 Leading white space used to be stripped from $spam_report which
2270 wrecked the formatting. Now it is preserved.
2272 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2273 that they are available at delivery time.
2275 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2277 TF/08 TLS error reporting now respects the incoming_interface and
2278 incoming_port log selectors.
2280 TF/09 Produce a more useful error message if an SMTP transport's hosts
2281 setting expands to an empty string.
2283 NM/06 Bugzilla 744: EXPN did not work under TLS.
2284 Patch provided by Phil Pennock.
2286 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2287 Patch provided by Richard Godbee.
2289 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2290 acl_smtp_notquit, added index entry.
2292 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2293 Patch provided by Eugene Bujak.
2295 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2296 accept(). Patch provided by Maxim Dounin.
2298 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2299 Patch provided by Phil Pennock.
2301 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2303 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2304 Patch provided by Brad "anomie" Jorsch.
2306 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2307 Patch provided by Dean Brooks.
2309 TK/05 Add native DKIM support (does not depend on external libraries).
2311 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2312 Patch provided by Graeme Fowler.
2314 NM/16 Bugzilla 851: Documentation example syntax fix.
2316 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2318 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2321 NM/19 Bugzilla 745: TLS version reporting.
2322 Patch provided by Phil Pennock.
2324 NM/20 Bugzilla 167: bool: condition support.
2325 Patch provided by Phil Pennock.
2327 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2328 clients. Patch provided by Phil Pennock.
2330 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2331 Patch provided by Brad "anomie" Jorsch.
2333 NM/23 Bugzilla 687: Fix misparses in eximstats.
2334 Patch provided by Heiko Schlittermann.
2336 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2337 Patch provided by Heiko Schlittermann.
2339 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2340 plus update to original patch.
2342 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2344 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2345 Patch provided by David Brownlee.
2347 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2349 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2351 NM/30 Bugzilla 888: TLS documentation bugfixes.
2353 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2355 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2356 Unlike the original bugzilla I have changed all shell scripts in src tree.
2358 NM/33 Bugzilla 898: Transport filter timeout fix.
2359 Patch by Todd Rinaldo.
2361 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2362 Patch by Serge Demonchaux.
2364 NM/35 Bugzilla 39: Base64 decode bug fixes.
2365 Patch by Jakob Hirsch.
2367 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2369 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2371 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2373 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2379 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2380 ALT-N's libdkim that I have put here:
2381 http://duncanthrax.net/exim-experimental/
2383 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2384 were called 'true' and it seems that C99 defines that as a reserved
2385 keyword to be used with 'bool' variable types. That means you could
2386 not include C99-style headers which use bools without triggering
2387 build errors in sieve.c.
2389 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2390 as mailq or other aliases. Changed the --help handling significantly
2391 to do whats expected. exim_usage() emits usage/help information.
2393 SC/01 Added the -bylocaldomain option to eximstats.
2395 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2397 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2399 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2405 PH/01 Another patch from the Sieve maintainer.
2407 PH/02 When an IPv6 address is converted to a string for single-key lookup
2408 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2409 dots are used instead of colons so that keys in lsearch files need not
2410 contain colons. This was done some time before quoting was made available
2411 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2412 (notated using the quote facility) so as to distinguish them from IPv4
2413 keys. This meant that lookups for IP addresses in host lists did not work
2414 for iplsearch lookups.
2416 This has been fixed by arranging for IPv6 addresses to be expressed with
2417 colons if the lookup type is iplsearch. This is not incompatible, because
2418 previously such lookups could never work.
2420 The situation is now rather anomalous, since one *can* have colons in
2421 ordinary lsearch keys. However, making the change in all cases is
2422 incompatible and would probably break a number of configurations.
2424 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2427 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2428 conversion specification without a maximum field width, thereby enabling
2429 a rogue spamd server to cause a buffer overflow. While nobody in their
2430 right mind would setup Exim to query an untrusted spamd server, an
2431 attacker that gains access to a server running spamd could potentially
2432 exploit this vulnerability to run arbitrary code as the Exim user.
2434 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2435 $primary_hostname instead of what libspf2 thinks the hosts name is.
2437 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2438 a directory entry by the name of the lookup key. Previously, if a
2439 symlink pointed to a non-existing file or a file in a directory that
2440 Exim lacked permissions to read, a lookup for a key matching that
2441 symlink would fail. Now it is enough that a matching directory entry
2442 exists, symlink or not. (Bugzilla 503.)
2444 PH/03 The body_linecount and body_zerocount variables are now exported in the
2447 PH/04 Added the $dnslist_matched variable.
2449 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2450 This means they are set thereafter only if the connection becomes
2453 PH/06 Added the client_condition to authenticators so that some can be skipped
2454 by clients under certain conditions.
2456 PH/07 The error message for a badly-placed control=no_multiline_responses left
2457 "_responses" off the end of the name.
2459 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2461 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2462 (without spaces) instead of just copying the configuration text.
2464 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2466 PH/11 Added $max_received_linelength.
2468 PH/12 Added +ignore_defer and +include_defer to host lists.
2470 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2471 way in which PCRE > 7.0 is built.
2473 PH/14 Implemented queue_only_load_latch.
2475 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2476 MAIL command. The effect was to mangle the value on 64-bit systems.
2478 PH/16 Another patch from the Sieve maintainer.
2480 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2482 PH/18 If a system quota error occurred while trying to create the file for
2483 a maildir delivery, the message "Mailbox is full" was not appended to the
2484 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2485 only to a quota excession during the actual writing of the file.
2487 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2488 characters?) which causes problems in log lines. The DN values are now
2489 passed through string_printing() before being added to log lines.
2491 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2492 and InterBase are left for another time.)
2494 PH/21 Added message_body_newlines option.
2496 PH/22 Guard against possible overflow in moan_check_errorcopy().
2498 PH/23 POSIX allows open() to be a macro; guard against that.
2500 PH/24 If the recipient of an error message contained an @ in the local part
2501 (suitably quoted, of course), incorrect values were put in $domain and
2502 $local_part during the evaluation of errors_copy.
2508 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2509 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2512 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2513 issue a MAIL command.
2515 PH/02 In an ACL statement such as
2517 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2519 if a client was not listed at all, or was listed with a value other than
2520 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2521 the condition was not true (as it should be), so access was not denied.
2522 The bug was that the ! inversion was incorrectly passed on to the second
2523 item. This has been fixed.
2525 PH/03 Added additional dnslists conditions == and =& which are different from
2526 = and & when the dns lookup returns more than one IP address.
2528 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2529 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2531 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2532 FSYNC, which compiles an option called disable_fsync that allows for
2533 bypassing fsync(). The documentation is heavily laced with warnings.
2535 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2537 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2538 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2539 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2540 including adding "make clean"; (3) Added -fPIC when compiling the test
2541 dynamically loaded module, to get rid of a warning.
2543 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2544 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2545 = 0s. The bug is otherwise harmless.
2547 PH/07 There was a bug in the dovecot authenticator such that the value of
2548 $auth1 could be overwritten, and so not correctly preserved, after a
2549 successful authentication. This usually meant that the value preserved by
2550 the server_setid option was incorrect.
2552 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2554 PH/09 Installed PCRE release 7.0.
2556 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2557 run for batched SMTP input. It is now run at the start of every message
2558 in the batch. While fixing this I discovered that the process information
2559 (output by running exiwhat) was not always getting set for -bs and -bS
2560 input. This is fixed, and it now also says "batched" for BSMTP.
2562 PH/11 Added control=no_pipelining.
2564 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2565 patch, slightly modified), and move the expansion of helo_data till after
2566 the connection is made in the smtp transport (so it can use these
2569 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2571 PH/14 Added log_selector = +pid.
2573 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2575 PH/16 Add ${if forany and ${if forall.
2577 PH/17 Added dsn_from option to vary the From: line in DSNs.
2579 PH/18 Flush SMTP output before performing a callout, unless control =
2580 no_callout_flush is set.
2582 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2583 was true (the default) a successful delivery failed to delete the retry
2584 item, thus causing premature timeout of the address. The bug is now
2587 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2589 PH/21 Long custom messages for fakedefer and fakereject are now split up
2590 into multiline responses in the same way that messages for "deny" and
2591 other ACL rejections are.
2593 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2594 with slight modification.
2596 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2597 draft, changing the syntax and factoring some duplicate code".
2599 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2600 for deliveries of the second and subsequent messages over the same SMTP
2603 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2604 ${reduce, with only minor "tidies".
2606 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2608 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2609 expansion side effects.
2611 PH/27 When a message times out after an over-quota error from an Exim-imposed
2612 quota, the bounce message says "mailbox is full". This message was not
2613 being given when it was a system quota that was exceeded. It now should
2616 MH/03 Made $recipients available in local_scan(). local_scan() already has
2617 better access to the recipient list through recipients_list[], but
2618 $recipients can be useful in postmaster-provided expansion strings.
2620 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2621 in the case of a MAIL command with additional options following the
2622 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2623 were accidentally chopped off.
2625 PH/29 SMTP synchronization checks are implemented when a command is read -
2626 there is a check that no more input is waiting when there shouldn't be
2627 any. However, for some commands, a delay in an ACL can mean that it is
2628 some time before the response is written. In this time, more input might
2629 arrive, invalidly. So now there are extra checks after an ACL has run for
2630 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2631 pipelining has not been advertised.
2633 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2635 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2636 returned in the SMTP response when smtp_return_error_details was false.
2637 This has been fixed.
2639 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2640 instead of the C I/O that was originally supplied, because problems were
2641 reported on Solaris.
2643 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2644 Exim which did not show up earlier: it was assuming that a call to
2645 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2646 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2647 was a macro that became an assignment, so it seemed to work. This has
2648 changed to a proper function call with a void return, hence the compile
2649 error. Exim's code has been fixed.
2651 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2654 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2656 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2658 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2659 because it turns out to be too big or there is a timeout) there is no
2660 "Completed" line in the log. When some messages of this type were
2661 selected by exigrep, they were listed as "not completed". Others were
2662 picked up by some special patterns. I have improved the selection
2663 criteria to be more general.
2665 PH/38 The host_find_failed option in the manualroute router can now be set
2666 to "ignore", to completely ignore a host whose IP address cannot be
2667 found. If all hosts are ignored, the behaviour is controlled by the new
2668 host_all_ignored option.
2670 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2671 homing or because of multiple MX records with /mx) generated more than
2672 one IP address, and the following item turned out to be the local host,
2673 all the secondary addresses of the first item were incorrectly removed
2674 from the list, along with the local host and any following hosts (which
2675 is what is supposed to happen).
2677 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2678 whoever called Exim into the -H file. In the case of the daemon it was
2679 behaving confusingly. When first started, it used values for whoever
2680 started the daemon, but after a SIGHUP it used the Exim user (because it
2681 calls itself on a restart). I have changed the code so that it now always
2684 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2685 message are rejected with the same error (e.g. no authentication or bad
2686 sender address), and a DATA command is nevertheless sent (as can happen
2687 with PIPELINING or a stupid MUA), the error message that was given to the
2688 RCPT commands is included in the rejection of the DATA command. This is
2689 intended to be helpful for MUAs that show only the final error to their
2692 PH/42 Another patch from the Sieve maintainer.
2694 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2695 Eximstats - Fixed some broken HTML links and added missing column headers
2697 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2698 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2700 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2702 PH/43 Yet another patch from the Sieve maintainer.
2704 PH/44 I found a way to check for a TCP/IP connection going away before sending
2705 the response to the final '.' that terminates a message, but only in the
2706 case where the client has not sent further data following the '.'
2707 (unfortunately, this is allowed). However, in many cases there won't be
2708 any further data because there won't be any more messages to send. A call
2709 to select() can be used: if it shows that the input is "ready", there is
2710 either input waiting, or the socket has been closed. An attempt to read
2711 the next input character can distinguish the two cases. Previously, Exim
2712 would have sent an OK response which the client would never have see.
2713 This could lead to message repetition. This fix should cure that, at
2714 least in a lot of common cases.
2716 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2717 advertised in response to EHLO.
2723 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2724 fixed by 4.65/MH/01 (is this a record?) are fixed:
2726 (i) An empty string was always treated as zero by the numeric comparison
2727 operators. This behaviour has been restored.
2729 (ii) It is documented that the numeric comparison operators always treat
2730 their arguments as decimal numbers. This was broken in that numbers
2731 starting with 0 were being interpreted as octal.
2733 While fixing these problems I realized that there was another issue that
2734 hadn't been noticed. Values of message_size_limit (both the global option
2735 and the transport option) were treated as octal if they started with 0.
2736 The documentation was vague. These values are now always treated as
2737 decimal, and I will make that clear in the documentation.
2743 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2744 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2747 MH/01 Don't check that the operands of numeric comparison operators are
2748 integers when their expansion is in "skipping" mode (fixes bug
2749 introduced by 4.64-PH/07).
2751 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2752 child addresses, Exim now panics and dies. Previously, because the count
2753 is held in a short int, deliveries were likely to be lost. As such a
2754 large number of recipients for a single message is ridiculous
2755 (performance will be very, very poor), I have chosen to impose a limit
2756 rather than extend the field.
2762 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2763 leftover -K file (the existence of which was triggered by #402).
2764 While we were at it, introduced process PID as part of the -K
2765 filename. This should rule out race conditions when creating
2768 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2769 processing considerably. Previous code took too long for large mails,
2770 triggering a timeout which in turn triggers #401.
2772 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2773 in the DK code in transports.c. sendfile() is not really portable,
2774 hence the _LINUX specificness.
2776 TF/01 In the add_headers option to the mail command in an Exim filter,
2777 there was a bug that Exim would claim a syntax error in any
2778 header after the first one which had an odd number of characters
2781 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2782 callout verification, Exim cached a "reject" for the entire domain. This
2783 is correct for most verifications, but it is not correct for a recipient
2784 verification with use_sender or use_postmaster set, because in that case
2785 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2786 case of MAIL FROM:<> rejection from other early rejections (e.g.
2787 rejection of HELO). When verifying a recipient using a non-null MAIL
2788 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2789 Whatever the result of the callout, the value of the domain cache is
2790 left unchanged (for any other kind of callout, getting as far as trying
2791 RCPT means that the domain itself is ok).
2793 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2796 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2797 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2798 session, the daemon ignores ECONNECT errors and logs others; it now
2799 ignores EPIPE as well.
2801 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2802 (quoted-printable decoding).
2804 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2805 later the small subsequent patch to fix an introduced bug.
2807 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2809 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2811 PH/08 An error is now given if message_size_limit is specified negative.
2813 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2814 to be given (somewhat) arbitrary names.
2816 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2819 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2820 miscellaneous code fixes
2822 PH/10 Added the log_reject_target ACL modifier to specify where to log
2825 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2826 hostname. This is wrong, because it relates to the incoming message (and
2827 probably the interface on which it is arriving) and not to the outgoing
2828 callout (which could be using a different interface). This has been
2829 changed to use the value of the helo_data option from the smtp transport
2830 instead - this is what is used when a message is actually being sent. If
2831 there is no remote transport (possible with a router that sets up host
2832 addresses), $smtp_active_hostname is used.
2834 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2835 tweaks were necessary in order to get it to work (see also 21 below):
2836 (a) The code assumed that strncpy() returns a negative number on buffer
2837 overflow, which isn't the case. Replaced with Exim's string_format()
2839 (b) There were several signed/unsigned issues. I just did the minimum
2840 hacking in of casts. There is scope for a larger refactoring.
2841 (c) The code used strcasecmp() which is not a standard C function.
2842 Replaced with Exim's strcmpic() function.
2843 (d) The code set only $1; it now sets $auth1 as well.
2844 (e) A simple test gave the error "authentication client didn't specify
2845 service in request". It would seem that Dovecot has changed its
2846 interface. Fortunately there's a specification; I followed it and
2847 changed what the client sends and it appears to be working now.
2849 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2852 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2853 address A is aliased to B and C, where B exists and C does not. Without
2854 -v the output is "A verified" because verification stops after a
2855 successful redirection if more than one address is generated. However,
2856 with -v the child addresses are also verified. Exim was outputting "A
2857 failed to verify" and then showing the successful verification for C,
2858 with its parentage. It now outputs "B failed to verify", showing B's
2859 parentage before showing the successful verification of C.
2861 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2862 look up a TXT record in a specific list after matching in a combined
2865 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2866 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2867 they consult the DNS. I had assumed they would set it the way they
2868 wanted; and indeed my experiments on Linux seem to show that in some
2869 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2870 To be on the safe side, however, I have now made the interface to
2871 host_find_byname() similar to host_find_bydns(), with an argument
2872 containing the DNS resolver options. The host_find_byname() function now
2873 sets these options at its start, just as host_find_bydns() does. The smtp
2874 transport options dns_qualify_single and dns_search_parents are passed to
2875 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2876 of host_find_byname() use the default settings of RES_DEFNAMES
2877 (qualify_single) but not RES_DNSRCH (search_parents).
2879 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2880 spool_read_header() do less string testing, by means of a preliminary
2881 switch on the second character of optional "-foo" lines. (This is
2882 overdue, caused by the large number of possibilities that now exist.
2883 Originally there were few.) While I was there, I also converted the
2884 str(n)cmp tests so they don't re-test the leading "-" and the first
2885 character, in the hope this might squeeze out yet more improvement.
2887 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2888 flag allowing group syntax was set by the header_syntax check but not
2889 turned off, possible causing trouble later; (2) The flag was not being
2890 set at all for the header_verify test, causing "group"-style headers to
2891 be rejected. I have now set it in this case, and also caused header_
2892 verify to ignore an empty address taken from a group. While doing this, I
2893 came across some other cases where the code for allowing group syntax
2894 while scanning a header line wasn't quite right (mostly, not resetting
2895 the flag correctly in the right place). These bugs could have caused
2896 trouble for malformed header lines. I hope it is now all correct.
2898 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2899 with the "reply" argument non-NULL. The code, however (which originally
2900 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2901 but it didn't always do it. This confused somebody who was copying the
2902 code for some other use. I have removed all the tests.
2904 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2905 feature that was used to support insecure browsers during the U.S. crypto
2906 embargo. It requires special client support, and Exim is probably the
2907 only MTA that supported it -- and would never use it because real RSA is
2908 always available. This code has been removed, because it had the bad
2909 effect of slowing Exim down by computing (never used) parameters for the
2910 RSA_EXPORT functionality.
2912 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2913 authenticator to fail if there's a tab character in the incoming data
2914 (there should never be unless someone is messing about, as it's supposed
2915 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2916 if the connection is using TLS or if the remote IP is the same as the
2917 local IP, and the "valid-client-cert option" if a client certificate has
2920 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2921 authenticators. This can be used for authorization after authentication
2922 succeeds. (In the case of plaintext, it servers for both authentication
2925 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2926 if any retry times were supplied.
2928 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2929 connection, where there is no client IP address to check. In this
2930 situation, the verify now always succeeds.
2932 PH/25 Applied John Jetmore's -Mset patch.
2934 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2936 PH/27 In a string expansion for a processed (not raw) header when multiple
2937 headers of the same name were present, leading whitespace was being
2938 removed from all of them, but trailing whitespace was being removed only
2939 from the last one. Now trailing whitespace is removed from each header
2940 before concatenation. Completely empty headers in a concatenation (as
2941 before) are ignored.
2943 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2944 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2946 PH/29 [Removed. This was a change that I later backed out, and forgot to
2947 correct the ChangeLog entry (that I had efficiently created) before
2948 committing the later change.]
2950 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2951 address errors (4xx response to RCPT) over the same connection as other
2952 messages routed to the same hosts. Such deliveries are always "forced",
2953 so retry times are not inspected. This resulted in far too many retries
2954 for the affected addresses. The effect occurred only when there were more
2955 hosts than the hosts_max_try setting in the smtp transport when it had
2956 the 4xx errors. Those hosts that it had tried were not added to the list
2957 of hosts for which the message was waiting, so if all were tried, there
2958 was no problem. Two fixes have been applied:
2960 (i) If there are any address or message errors in an SMTP delivery, none
2961 of the hosts (tried or untried) are now added to the list of hosts
2962 for which the message is waiting, so the message should not be a
2963 candidate for sending over the same connection that was used for a
2964 successful delivery of some other message. This seems entirely
2965 reasonable: after all the message is NOT "waiting for some host".
2966 This is so "obvious" that I'm not sure why it wasn't done
2967 previously. Hope I haven't missed anything, but it can't do any
2968 harm, as the worst effect is to miss an optimization.
2970 (ii) If, despite (i), such a delivery is accidentally attempted, the
2971 routing retry time is respected, so at least it doesn't keep
2972 hammering the server.
2974 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2975 in ${readsocket because some servers need this prod.
2977 PH/32 Added some extra debug output when updating a wait-xxx database.
2979 PH/33 The hint "could be header name not terminated by colon", which has been
2980 given for certain expansion errors for a long time, was not being given
2981 for the ${if def:h_colon_omitted{... case.
2983 PH/34 The spec says: "With one important exception, whenever a domain list is
2984 being scanned, $domain contains the subject domain." There was at least
2985 one case where this was not true.
2987 PH/35 The error "getsockname() failed: connection reset by peer" was being
2988 written to the panic log as well as the main log, but it isn't really
2989 panic-worthy as it just means the connection died rather early on. I have
2990 removed the panic log writing for the ECONNRESET error when getsockname()
2993 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2994 runs only) independently of the message's sender address. This meant
2995 that, if the 4xx error was in fact related to the sender, a different
2996 message to the same recipient with a different sender could confuse
2997 things. In particular, this can happen when sending to a greylisting
2998 server, but other circumstances could also provoke similar problems.
2999 I have changed the default so that the retry time for these errors is now
3000 based a combination of the sender and recipient addresses. This change
3001 can be overridden by setting address_retry_include_sender=false in the
3004 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3005 remote server are returned as part of bounce messages. This was not
3006 happening for LMTP over a pipe (the lmtp transport), but now it is the
3007 same for both kinds of LMTP.
3009 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3010 in header lines that were in fact CNAMEs. This is no longer the case.
3012 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3013 and queue runs started by the daemon processed all messages. This has
3014 been fixed so that -R and -S can now usefully be given with -q<time>.
3016 PH/40 Import PCRE release 6.7 (fixes some bugs).
3018 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3020 PH/42 Give an error if -q is specified more than once.
3022 PH/43 Renamed the variables $interface_address and $interface_port as
3023 $received_ip_address and $received_port, to make it clear that these
3024 values apply to message reception, and not to the outgoing interface when
3025 a message is delivered. (The old names remain recognized, of course.)
3027 PH/44 There was no timeout on the connect() call when using a Unix domain
3028 socket in the ${readsocket expansion. There now is.
3030 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3031 be meaningful with "accept".
3033 SC/01 Eximstats V1.43
3034 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3036 SC/02 Eximstats V1.44
3037 Use a glob alias rather than an array ref in the generated
3038 parser. This improves both readability and performance.
3040 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3041 Collect SpamAssassin and rejection statistics.
3042 Don't display local sender or destination tables unless
3043 there is data to show.
3044 Added average volumes into the top table text output.
3046 SC/04 Eximstats V1.46
3047 Collect data on the number of addresses (recipients)
3048 as well as the number of messages.
3050 SC/05 Eximstats V1.47
3051 Added 'Message too big' to the list of mail rejection
3052 reasons (thanks to Marco Gaiarin).
3054 SC/06 Eximstats V1.48
3055 Mainlog lines which have GMT offsets and are too short to
3056 have a flag are now skipped.
3058 SC/07 Eximstats V1.49 (Alain Williams)
3059 Added the -emptyok flag.
3061 SC/08 Eximstats V1.50
3062 Fixes for obtaining the IP address from reject messages.
3064 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3065 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3066 whitespace changes from 4.64-PH/27
3068 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3071 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3072 are found, allow negative numbers in numeric criteria)
3074 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3076 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3077 to match changes made in 4.64-PH/43
3079 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3081 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3082 "domain" is actually an IP address, and force a failure. This locks out
3083 those revolvers/nameservers that support "A-for-A" lookups, in
3084 contravention of the specifications.
3086 PH/48 When a host name was looked up from an IP address, and the subsequent
3087 forward lookup of the name timed out, the host name was left in
3088 $sender_host_name, contrary to the specification.
3090 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3091 restricted to single-key lookups, Exim was not diagnosing an error if
3092 * or *@ was used with a query-style lookup.
3094 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3096 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3097 long ago, but noone interested enough thought of it. Let's just say that
3098 the "1.1" means that there are some new functions that weren't there at
3099 some point in the past.
3101 PH/51 Error processing for expansion failure of helo_data from an smtp
3102 transport during callout processing was broken.
3104 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3105 tested/used via the -bh/-bhc/-bs options.
3107 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3108 bug, fixed in subsequent PCRE releases).
3110 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3111 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3113 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3119 SC/01 Use a glob alias rather than an array ref in eximstats generated
3120 parser. This improves both readability and performance.
3122 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3123 Don't display local sender or destination tables in eximstats unless
3124 there is data to show.
3125 Added average volumes into the eximstats top table text output.
3127 SC/03 Collect data on the number of addresses (recipients) as well
3128 as the number of messages in eximstats.
3130 TF/01 Correct an error in the documentation for the redirect router. Exim
3131 does (usually) call initgroups() when daemonizing.
3133 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3134 with consistent privilege compared to when running as a daemon.
3136 TF/03 Note in the spec that $authenticated_id is not set for local
3137 submissions from trusted users.
3139 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3140 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3142 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3143 by adding some example configuration directives to the default
3144 configuration file. A little bit of work is required to uncomment the
3145 directives and define how usernames and passwords are checked, but
3146 there is now a framework to start from.
3148 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3149 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3150 without this. I don't know how relevant this is to other LDAP libraries.
3152 PH/02 Add the verb name to the "unknown ACL verb" error.
3154 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3156 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3158 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3159 directory not expanded when it should be if an expanded home directory
3160 was set for the address (which is overridden by the transport).
3162 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3165 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3166 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3167 because it is too late at that time, and has no effect.
3169 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3170 security issue with \' (bugzilla #107). I could not use the
3171 PQescapeStringConn() function, because it needs a PGconn value as one of
3174 PH/08 When testing addresses using -bt, indicate those final addresses that
3175 are duplicates that would not cause an additional delivery. At least one
3176 person was confused, thinking that -bt output corresponded to deliveries.
3177 (Suppressing duplicates isn't a good idea as you lose the information
3178 about possibly different redirections that led to the duplicates.)
3180 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3181 systems where poll() doesn't work, in particular OS X.
3183 PH/10 Added more information to debugging output for retry time not reached.
3185 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3186 operations in malware.c.
3188 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3191 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3192 syslog_duplication set false, log lines that would normally be written
3193 both the the main log and to the reject log were not written to syslog at
3196 PH/14 In the default configuration, change the use of "message" in ACL warn
3197 statements to "add_header".
3199 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3200 not followed by a command (e.g. "seen endif").
3202 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3203 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3206 PH/17 Added extra conditions to the default value of delay_warning_condition
3210 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3211 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3212 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3215 The Auto-Submitted: and various List- headers are standardised, whereas I
3216 don't think Precedence: ever was.
3218 PH/18 Refactored debugging code in route_finduser() to show more information,
3219 in particular, the error code if getpwnam() issues one.
3221 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3222 This is apparently needed in addition to the PH/07 change above to avoid
3223 any possible encoding problems.
3225 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3226 but not after initializing Perl.
3228 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3229 output them only if debugging. By default they are written stderr,
3230 apparently, which is not desirable.
3232 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3235 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3238 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3240 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3241 authenticated or an ident call has been made. Suppress the default
3242 values for $authenticated_id and $authenticated_sender (but permit -oMai
3243 and -oMas) when testing with -bh.
3245 PH/24 Re-jigged the order of the tests in the default configuration so that the
3246 tests for valid domains and recipients precede the DNS black list and CSA
3247 tests, on the grounds that those ones are more expensive.
3249 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3250 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3251 This bug exists in every version of Exim that I still have, right back to
3254 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3255 However, an attempt to turn on case-sensitivity in a regex key by
3256 including (?-i) didn't work because the subject string was already
3257 lowercased, and the effects were non-intuitive. It turns out that a
3258 one-line patch can be used to allow (?-i) to work as expected.
3264 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3265 other effects) broke the use of negated acl sub-conditions.
3267 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3270 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3271 "Deny" causes Exim to reject the incoming connection with a 554 error.
3272 Unfortunately, if there is a major crisis, such as a disk failure,
3273 tcp-wrappers gives "deny", whereas what one would like would be some
3274 kind of temporary error. A kludge has been added to help with this.
3275 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3276 554 error is used if errno is still zero or contains ENOENT (which occurs
3277 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3280 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3282 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3283 errors. Otherwise a message that provokes a temporary error (when other
3284 messages do not) can cause a whole host to time out.
3286 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3287 addresses were routed directly to files or pipes from a redirect router.
3288 File deliveries just didn't batch; pipe deliveries might have suffered
3291 PH/06 A failure to get a lock for a hints database would erroneously always say
3292 "Failed to get write lock", even when it was really a read lock.
3294 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3295 of 0600. This has been changed to use the value of the lockfile_mode
3296 option (which defaults to 0600).
3298 PH/08 Applied small patch from the Sieve maintainer.
3300 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3301 folder from quota calculations, a direct delivery into this folder messed
3302 up the contents of the maildirsize file. This was because the regex was
3303 used only to exclude .Trash (or whatever) when the size of the mailbox
3304 was calculated. There was no check that a delivery was happening into an
3305 excluded directory. This bug has been fixed by ignoring all quota
3306 processing for deliveries into excluded directories.
3308 PH/10 Added the maildirfolder_create_regex option to appendfile.
3314 PH/01 The code for finding all the local interface addresses on a FreeBSD
3315 system running IPv6 was broken. This may well have applied to all BSD
3316 systems, as well as to others that have similar system calls. The broken
3317 code found IPv4 interfaces correctly, but gave incorrect values for the
3318 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3319 that it would not match correctly against @[] and not recognize the IPv6
3322 PH/02 The ipliteral router was not recognizing addresses of the form user@
3323 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3325 PH/03 Added disable_ipv6.
3327 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3328 decoded form, because it is most often used to construct To: headers
3329 lines in autoreplies, and the decoded form may well be syntactically
3330 invalid. However, $reply_address has leading white space removed, and all
3331 newlines turned into spaces so that the autoreply transport does not
3334 PH/05 If group was specified without a user on a router, and no group or user
3335 was specified on a transport, the group from the router was ignored.
3337 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3338 for visible compile-time settings that can be used to change these
3339 numbers, for those that want even more. Backwards compatibility with old
3340 spool files has been maintained. However, going back to a previous Exim
3341 release will lost any variables that are in spool files.
3343 PH/07 Two small changes when running in the test harness: increase delay when
3344 passing a TCP/IP connection to a new process, in case the original
3345 process has to generate a bounce, and remove special handling of
3346 127.0.0.2 (sic), which is no longer necessary.
3348 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3349 be the same on different OS.
3351 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3354 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3355 whether --show-vars was specified or not
3357 JJ/02 exipick: Added support for new ACL variable spool format introduced
3360 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3361 syntactically invalid From: or Reply-to: line, and a filter used this to
3362 generate an autoreply, and therefore failed to obtain an address for the
3363 autoreply, Exim could try to deliver to a non-existent relative file
3364 name, causing unrelated and misleading errors. What now happens is that
3365 it logs this as a hard delivery error, but does not attempt to create a
3368 PH/11 The exinext utility has a -C option for testing purposes, but although
3369 the given file was scanned by exinext itself; it wasn't being passed on
3370 when Exim was called.
3372 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3373 an end-of-file indication when reading a command response.
3375 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3376 compiled. In many other places in Exim, IPv6 addresses are always
3377 recognized, so I have changed this. It also means that IPv4 domain
3378 literals of the form [IPV4:n.n.n.n] are now always recognized.
3380 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3381 used if the router is not running as root, for example, when verifying at
3382 ACL time, or when using -bh. The debugging output from this situation was
3383 non-existent - all you got was a failure to exec. I have made two
3386 (a) Failures to set uid/gid, the current directory, or a process leader
3387 in a subprocess such as that created by queryprogram now generate
3388 suitable debugging output when -d is set.
3390 (b) The queryprogram router detects when it is not running as root,
3391 outputs suitable debugging information if -d is set, and then runs
3392 the subprocess without attempting to change uid/gid.
3394 PH/15 Minor change to Makefile for building test_host (undocumented testing
3397 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3398 additional section of a DNS packet that returns MX or SRV records.
3399 Instead, it always explicitly searches for A/AAAA records. This avoids
3400 major problems that occur when a DNS server includes only records of one
3401 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3402 fixed another bug: if SRV records were looked up and the corresponding
3403 address records were *not* found in the additional section, the port
3404 values from the SRV records were lost.
3406 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3407 using the correct key (the original address) when searching the retry
3408 rules in order to find which one to use for generating the retry hint.
3410 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3411 adding the default one. Similarly, if it contains a Reply-To: header, the
3412 errors_reply_to option, if set, is not used.
3414 PH/19 When calculating a retry time, Exim used to measure the "time since
3415 failure" by looking at the "first failed" field in the retry record. Now
3416 it does not use this if it is later than than the arrival time of the
3417 message. Instead it uses the arrival time. This makes for better
3418 behaviour in cases where some deliveries succeed, thus re-setting the
3419 "first failed" field. An example is a quota failure for a huge message
3420 when small messages continue to be delivered. Without this change, the
3421 "time since failure" will always be short, possible causing more frequent
3422 delivery attempts for the huge message than are intended.
3423 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3425 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3426 $1, $2, $3) because the numerical variables can be reset during some
3427 expansion items (e.g. "match"), thereby losing the authentication data.
3429 PH/21 Make -bV show the size of off_t variables so that the test suite can
3430 decide whether to run tests for quotas > 2G.
3432 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3433 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3434 filecount value is greater than 2G or if a quota value is greater than 2G
3435 on a system where the size of off_t is not greater than 4, a panic error
3438 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3439 never match. The debug and -bh output now contains an explicit error
3440 message indicating a malformed IPv4 address or mask.
3442 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3443 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3444 PH/23 above applies.
3446 PH/25 Do not write to syslog when running in the test harness. The only
3447 occasion when this arises is a failure to open the main or panic logs
3448 (for which there is an explicit test).
3450 PH/26 Added the /no_tell option to "control=freeze".
3452 PH/27 If a host name lookup failed very early in a connection, for example, if
3453 the IP address matched host_lookup and the reverse lookup yielded a name
3454 that did not have a forward lookup, an error message of the form "no IP
3455 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3456 could be logged. Now it outputs the IP address instead of "NULL".
3458 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3459 allows the sender and the authenticated sender to be set when
3460 submitting a message from within Exim. Since child_open_exim() is
3461 documented for local_scan(), the new function should be too.
3463 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3464 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3465 results in an empty string is now treated as unset.
3467 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3469 PH/31 Added sender_verify_fail logging option.
3471 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3472 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3473 tidied the source and removed it altogether.
3475 PH/33 When a queue run was abandoned because the load average was too high, a
3476 log line was always written; now it is written only if the queue_run log
3477 selector is set. In addition, the log line for abandonment now contains
3478 information about the queue run such as the pid. This is always present
3479 in "start" and "stop" lines but was omitted from the "abandon" line.
3481 PH/34 Omit spaces between a header name and the colon in the error message that
3482 is given when verify = headers_syntax fails (if there are lots of them,
3483 the message gets confusing).
3485 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3486 names, as there are now some PTR records that contain slashes. This check
3487 is only to protect against broken name servers that fall over on strange
3488 characters, so the fact that it applies to all lookups doesn't matter.
3490 PH/36 Now that the new test suite is complete, we can remove some of the
3491 special code in Exim that was needed for the old test suite. For example,
3492 sorting DNS records because real resolvers return them in an arbitrary
3493 order. The new test suite's fake resolver always returns records in the
3496 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3497 bounces) except when queue_only is set, to avoid logging races between
3498 the different processes.
3500 PH/38 Panic-die if .include specifies a non-absolute path.
3502 PH/39 A tweak to the "H" retry rule from its user.
3504 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3505 a label. They prevented compilation on older perls.
3507 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3508 a warning to be raised on newish perls.
3510 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3511 on queue. Changes to match documented behaviour of showing count of
3512 messages matching specified criteria.
3514 PH/40 Changed the default ident timeout from 30s to 5s.
3516 PH/41 Added support for the use of login_cap features, on those BSD systems
3517 that have them, for controlling the resources used by pipe deliveries.
3519 PH/42 The content-scanning code uses fopen() to create files in which to put
3520 message data. Previously it was not paying any attention to the mode of
3521 the files. Exim runs with umask(0) because the rest of the code creates
3522 files with open(), and sets the required mode explicitly. Thus, these
3523 files were ending up world-writeable. This was not a big issue, because,
3524 being within the spool directory, they were not world-accessible. I have
3525 created a function called modefopen, which takes an additional mode
3526 argument. It sets umask(777), creates the file, chmods it to the required
3527 mode, then resets the umask. All the relevant calls to fopen() in the
3528 content scanning code have been changed to use this function.
3530 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3531 to 24 hours. This avoids potential overflow problems when processing G
3532 and H retry rules. I suspect nobody ever tinkers with this value.
3534 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3536 PH/45 When the plaintext authenticator is running as a client, the server's
3537 challenges are checked to ensure they are valid base64 strings. By
3538 default, the authentication attempt is cancelled if an invalid string is
3539 received. Setting client_ignore_invalid_base64 true ignores these errors.
3540 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3541 they are received. Thus, the responses can be made to depend on the
3542 challenges. If an invalid string is ignored, an empty string is placed in
3545 PH/46 Messages that are created by the autoreply transport now contains a
3546 References: header, in accordance with RFCs 2822 and 3834.
3548 PH/47 Added authenticated_sender_force to the smtp transport.
3550 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3552 PH/49 Installed latest patch from the Sieve maintainer.
3554 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3555 was also set, the appendfile transport was unnecessarily scanning a
3556 directory of message files (e.g. for maildir delivery) to find the count
3557 of files (along with the size), even though it did not need this
3558 information. It now does the scan only if it needs to find either the
3559 size of the count of files.
3561 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3563 PH/52 Two bugs concerned with error handling when the smtp transport is
3566 (i) Exim was not creating retry information for temporary errors given
3567 for individual recipients after the DATA command when the smtp transport
3568 was used in LMTP mode. This meant that they could be retried too
3569 frequently, and not timed out correctly.
3571 (ii) Exim was setting the flag that allows error details to be returned
3572 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3573 recipients that were returned after the DATA command.
3575 PH/53 This is related to PH/52, but is more general: for any failing address,
3576 when detailed error information was permitted to be returned to the
3577 sender, but the error was temporary, then after the final timeout, only
3578 "retry timeout exceeded" was returned. Now it returns the full error as
3579 well as "retry timeout exceeded".
3581 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3582 do this, and (what is worse) MTAs that accept it.
3584 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3585 will now be deprecated.
3587 PH/56 New os.c-cygwin from the Cygwin maintainer.
3589 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3590 formats (previously only available in exim formats via -bpr, -bpru,
3591 and -bpra. Now also available in native and exiqgrep formats)
3593 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3594 with very large, slow to parse queues
3596 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3598 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3600 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3601 responses to authentication challenges, though it was showing the
3602 challenges; (ii) I've removed the CR characters from the debug output for
3605 PH/58 Allow for the insertion of a newline as well as a space when a string
3606 is turned into more than one encoded-word during RFC 2047 encoding. The
3607 Sieve code now uses this.
3609 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3610 data_4xx, lost_connection, tls_required.
3612 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3613 message was being sent as an SMTP response.
3615 PH/61 Add -l and -k options to exicyclog.
3617 PH/62 When verifying, if an address was redirected to one new address, so that
3618 verification continued, and the new address failed or deferred after
3619 having set something in $address_data, the value of $address_data was not
3620 passed back to the ACL. This was different to the case when no
3621 redirection occurred. The value is now passed back in both cases.
3623 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3624 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3625 use login_cap.h, so on its own it isn't the distinguishing feature. The
3626 new name refers directly to the setclassresources() function.
3628 PH/65 Added configuration files for NetBSD3.
3630 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3632 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3633 is preferred over IPv4.
3635 PH/68 The bounce_return_message and bounce_return_body options were not being
3636 honoured for bounces generated during the reception of non-SMTP messages.
3637 In particular, this applied to messages rejected by the ACL. This bug has
3638 been fixed. However, if bounce_return_message is true and bounce_return_
3639 body is false, the headers that are returned for a non-SMTP message
3640 include only those that have been read before the error was detected.
3641 (In the case of an ACL rejection, they have all been read.)
3643 PH/69 The HTML version of the specification is now built in a directory called
3644 spec_html instead of spec.html, because the latter looks like a path with
3645 a MIME-type, and this confuses some software.
3647 PH/70 Catch two compiler warnings in sieve.c.
3649 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3650 function verify_get_ident() calls ip_connect() to connect a socket, but
3651 if the "connect()" function timed out, ip_connect() used to close the
3652 socket. However, verify_get_ident() also closes the socket later, and in
3653 between Exim writes to the log, which may get opened at this point. When
3654 the socket was closed in ip_connect(), the log could get the same file
3655 descriptor number as the socket. This naturally causes chaos. The fix is
3656 not to close the socket in ip_connect(); the socket should be closed by
3657 the function that creates it. There was only one place in the code where
3658 this was missing, in the iplookup router, which I don't think anybody now
3659 uses, but I've fixed it anyway.
3661 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3662 well as to direct DNS lookups. Otherwise the handling of names in host
3663 lists is inconsistent and therefore confusing.
3669 PH/01 Two changes to the default runtime configuration:
3671 (1) Move the checks for relay_from_hosts and authenticated clients from
3672 after to before the (commented out) DNS black list checks.
3674 (2) Add control=submission to the relay_from_hosts and authenticated
3675 clients checks, on the grounds that messages accepted by these
3676 statements are most likely to be submissions.
3678 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3680 (1) Generate an error if the third argument for the ${prvs expansion is
3683 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3686 (3) Reset the variables that are obtained from the first argument of
3687 ${prvscheck and used in the second argument before leaving the code,
3688 because their memory is reclaimed, so using them afterwards may do
3691 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3692 one (it's much easier than Tom thought :-).
3694 (5) Because of (4), we can now allow for the use of $prvscheck_result
3695 inside the third argument.
3697 PH/03 For some reason, the default setting of PATH when running a command from
3698 a pipe transport was just "/usr/bin". I have changed it to
3701 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3702 anything to be listed in the output from -bV.
3704 PH/05 When a filter generated an autoreply, the entire To: header line was
3705 quoted in the delivery log line, like this:
3707 => >A.N.Other <ano@some.domain> <original@ddress> ...
3709 This has been changed so that it extracts the operative address. There
3710 may be more than one such address. If so, they are comma-separated, like
3713 => >ano@some.domain,ona@other.domain <original@ddress> ...
3715 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3716 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3717 not being looked up in the rDNS to get a host name, Exim was showing the
3718 IP address twice in Received: lines, even though the IP addresses were
3719 identical. For example:
3721 Received: from [1.2.3.4] (helo=[1.2.3.4])
3723 However, if the real host name was known, it was omitting the HELO data
3724 if it matched the actual IP address. This has been tidied up so that it
3725 doesn't show the same IP address twice.
3727 PH/07 When both +timestamp and +memory debugging was on, the value given by
3728 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3729 called by the debug printing, thereby overwriting the timestamp buffer.
3730 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3732 PH/08 When the original message was included in an autoreply transport, it
3733 always said "this is a copy of the message, including all the headers",
3734 even if body_only or headers_only was set. It now gives an appropriate
3737 PH/09 Applied a patch from the Sieve maintainer which:
3739 o fixes some comments
3740 o adds the (disabled) notify extension core
3741 o adds some debug output for the result of if/elsif tests
3742 o points to the current vacation draft in the documentation
3743 and documents the missing references header update
3747 o fixes a bug in processing the envelope test (when testing
3748 multiple envelope elements, the last element determined the
3751 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3752 Electronic Mail") by including:
3754 Auto-submitted: auto-generated
3756 in the messages that it generates (bounce messages and others, such as
3757 warnings). In the case of bounce messages for non-SMTP messages, there was
3758 also a typo: it was using "Auto_submitted" (underscore instead of
3759 hyphen). Since every message generated by Exim is necessarily in response
3760 to another message, thes have all been changed to:
3762 Auto-Submitted: auto-replied
3764 in accordance with these statements in the RFC:
3766 The auto-replied keyword:
3768 - SHOULD be used on messages sent in direct response to another
3769 message by an automatic process,
3771 - MUST NOT be used on manually-generated messages,
3773 - MAY be used on Delivery Status Notifications (DSNs) and Message
3774 Disposition Notifications (MDNs),
3776 - MUST NOT be used on messages generated by automatic or periodic
3777 processes, except for messages which are automatic responses to
3780 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3781 to the default Received: header definition.
3783 PH/12 Added log selector acl_warn_skipped (default on).
3785 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3786 variables because (a) they are in the wrong storage pool and (b) even if
3787 they were copied, it wouldn't work properly because of the caching.
3789 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3790 checking when decoding. Apparently there are clients that generate
3791 overlong encoded strings. Why am I not surprised?
3793 PH/15 If the first argument of "${if match_address" was not empty, but did not
3794 contain an "@" character, Exim crashed. Now it writes a panic log message
3795 and treats the condition as false.
3797 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3799 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3800 extension "envelope-auth". The code is finished and in agreement with
3801 other implementations, but there is no documentation so far and in fact,
3802 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3803 not changing the active code.
3805 Print executed "if" and "elsif" statements when debugging is used. This
3806 helps a great deal to understand what a filter does.
3808 Document more things not specified clearly in RFC3028. I had all this
3809 sorted out, when out of a sudden new issues came to my mind. Oops."
3811 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3814 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3815 canonical form (as documented). However, after a host name lookup from
3816 the IP address, check_host() was doing a simple string comparison with
3817 addresses acquired from the DNS when checking that the found name did
3818 have the original IP as one of its addresses. Since any found IPv6
3819 addresses are likely to be in abbreviated form, the comparison could
3820 fail. Luckily, there already exists a function for doing the comparison
3821 by converting both addresses to binary, so now that is used instead of
3822 the text comparison.
3824 PH/20 There was another similar case to PH/19, when a complete host name was
3825 given in a host list; looking up its IP address could give an abbreviated
3826 form, whereas the current host's name might or might not be abbreviated.
3827 The same fix has been applied.
3833 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3834 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3837 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3838 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3840 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3842 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3843 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3844 Also, for "Auto-Submitted", treat anything other than "no" as
3845 non-personal, in accordance with RFC 3834. (Previously it treated
3846 anything starting "auto-" as non-personal.)
3848 TF/01 The control=submission/name=... option had a problem with syntax
3849 errors if the name included a slash character. The /name= option
3850 now slurps the rest of the string, so it can include any characters
3851 but it must come last in the list of options (after /sender_retain
3854 PH/05 Some modifications to the interface to the fake nameserver for the new
3862 TK/01 Added the "success_on_redirect" address verification option. See
3863 NewStuff for rationale and an example.
3865 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3867 PH/02 Patch to exigrep to allow it to work on syslog lines.
3869 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3870 fread() to skip over the body file's header line, because in Cygwin the
3871 header line is locked and is inaccessible.
3873 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3874 co-exist for some time) to make it clear that it is the Exim ID that is
3875 referenced, not the Message-ID: header line.
3877 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3878 string_format() function, because snprintf() does not exist on all
3881 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3882 ${stat: expansion item.
3884 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3885 protocol synchronization error", to keep the pedants happy.
3887 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3888 well as for IRIX systems, when gcc is being used. See the host.c source
3891 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3893 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3896 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3897 transport and to the smtp transport in LMTP mode.
3899 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3901 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3902 run a filter in a subprocess. This could lead to confusion in subsequent
3903 lookups in the parent process. There should also be a search_tidyup() at
3904 the end of the subprocess.
3906 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3907 only if the host matched helo_try_verify_hosts, which caused the
3908 verification to occur when the EHLO/HELO command was issued. The ACL just
3909 tested the remembered result. Now, if a previous verification attempt has
3910 not happened, "verify = helo" does it there and then.
3912 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3914 TK/03 Fix log output including CR from clamd.
3916 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3917 exist provoked a memory error which could cause a segfault.
3919 PH/15 Installed PCRE 6.2
3921 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3923 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3924 of the problem. Specifically, suggested +O2 rather than +O1 for the
3927 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3929 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3930 record, the port was not correctly logged when the outgoing_port log
3931 selector was set (it logged the transort's default port).
3933 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3934 fallback_hosts, and "hosts" in the smtp transport.
3936 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3937 host errors such as "Connection refused".
3939 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3940 authentication with radiusclient 0.4.9:
3942 - Error returned from rc_read_config was caught wrongly
3943 - Username/password not passed on to radius server due to wrong length.
3945 The presumption is that some radiusclient API changes for 4.51/PH/17
3946 were not taken care of correctly. The code is still untested by me (my
3947 Linux distribution still has 0.3.2 of radiusclient), but it was
3948 contributed by a Radius user.
3950 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3951 expanding the "port" option of the smtp transport.
3953 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3954 while reading a MIME header. Thanks to Tom Hughes for a patch.
3956 PH/24 Include config.h inside local_scan.h so that configuration settings are
3959 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3960 that in an ACL for RCPT (for example), you can examine exactly what was
3963 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3964 commands, but it was not correctly comparing the address with the actual
3965 client host address. Thus, it would show the EHLO address in Received:
3966 header lines when this was not necessary.
3968 PH/27 Added the % operator to ${eval:}.
3970 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3971 it should be ignoring failures (because with -C, for example, it has lost
3972 privilege). It wasn't ignoring creation failures other than "already
3975 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3978 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3979 because an input line was too long, either on its own, or by virtue of
3980 too many continuations, the temporary file was not being removed, and the
3981 return code was incorrect.
3983 PH/31 Missing "BOOL" in function definition in filtertest.c.
3985 PH/32 Applied Sieve patches from the maintainer.
3987 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3989 PH/33 Added "verify = not_blind".
3991 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3992 Local/Makefile (with some defaults set). These are used in built scripts
3993 such as exicyclog, but they have never been used in the exim_install
3994 script (though there are many overriding facilities there). I have
3995 arranged that the exim_install script now takes note of these two
3998 PH/35 Installed configuration files for Dragonfly.
4000 PH/36 When a locally submitted message by a trusted user did not contain a
4001 From: header, and the sender address was obtained from -f or from an SMTP
4002 MAIL command, and the trusted user did not use -F to supply a sender
4003 name, $originator_name was incorrectly used when constructing a From:
4004 header. Furthermore, $originator_name was used for submission mode
4005 messages from external hosts without From: headers in a similar way,
4006 which is clearly wrong.
4008 PH/37 Added control=suppress_local_fixups.
4010 PH/38 When log_selector = +received_sender was set, and the addition of the
4011 sender made the log line's construction buffer exactly full, or one byte
4012 less than full, an overflow happened when the terminating "\n" was
4015 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4016 when the result of a list match is failure because a DNS lookup failed.
4018 PH/40 RM_COMMAND is now used in the building process.
4020 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4021 the "build-* directories that it finds.
4023 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4024 address in a domain literal was a prefix of an interface address.
4026 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4027 when verifying a sender address, unless rewrite_headers is false.
4029 PH/44 Wrote a long comment about why errors_to addresses are verified as
4030 recipients, not senders.
4032 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4033 the ratelimit ACL was added.
4035 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4037 PH/46 Added extra information about PostgreSQL errors to the error string.
4039 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4040 avoiding the need to install special zones in a real server. This is
4041 backwards compatible; if it can't find the fake resolver, it drops back.
4042 Thus, both old and new test suites can be run.
4044 TF/02 Added util/ratelimit.pl
4046 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4047 clock is set back in time.
4049 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4050 Candler <B.Candler@pobox.com>.
4052 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4053 OK for addresses that are the result of redirections.
4055 PH/48 A number of further additions for the benefit of the new test suite,
4056 including a fake gethostbyname() that interfaces to the fake DNS resolver
4059 TF/06 The fix for widen_domains has also been applied to qualify_single and
4060 search_parents which are the other dnslookup options that can cause
4063 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4066 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4068 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4069 service" if it encountered an MX record with an empty target hostname.
4070 The message is now "an MX or SRV record indicated no SMTP service".
4072 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4073 if the DNS of the sending site is misconfigured. This is quite a
4074 common situation. This change restores the behaviour of treating a
4075 helo verification defer as a failure.
4077 PH/51 If self=fail was set on a router, the bounce message did not include the
4078 actual error message.
4084 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4086 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4087 command itself ended in error, the underlying message about the transport
4088 filter timeout was being overwritten with the pipe command error. Now the
4089 underlying error message should be appended to the second error message.
4091 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4093 PH/02 Reduce the amount of output that "make" produces by default. Full output
4094 can still be requested.
4096 PH/03 The warning log line about a condition test deferring for a "warn" verb
4097 was being output only once per connection, rather than after each
4098 occurrence (because it was using the same function as for successful
4099 "warn" verbs). This seems wrong, so I have changed it.
4101 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4102 it should not have, which might have caused a crash in the right
4103 circumstances, but probably never did.
4105 PH/04 Installed a modified version of Tony Finch's patch to make submission
4106 mode fix the return path as well as the Sender: header line, and to
4107 add a /name= option so that you can make the user's friendly name appear
4110 TF/03 Added the control = fakedefer ACL modifier.
4112 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4113 Mark Lowes for thorough testing.
4115 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4117 TK/03 Merged latest SRS patch from Miles Wilton.
4119 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4120 with the definition in sysexits.h (which is #included earlier).
4121 Fortunately, Exim does not actually use EX_OK. The code used to try to
4122 preserve the sysexits.h value, by assuming that macro definitions were
4123 scanned for macro replacements. I have been disabused of this notion,
4124 so now the code just undefines EX_OK before #including unistd.h.
4126 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4127 in the smtp transport. When a block could not be written in a single
4128 write() function, the timeout was being re-applied to each part-write.
4129 This seems wrong - if the receiver was accepting one byte at a time it
4130 would take for ever. The timeout is now adjusted when this happens. It
4131 doesn't have to be particularly precise.
4133 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4134 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4136 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4137 without a domain if the check to <postmaster@domain> fails.
4139 SC/01 Eximstats: added -xls and the ability to specify output files
4140 (patch written by Frank Heydlauf).
4142 SC/02 Eximstats: use FileHandles for outputting results.
4144 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4146 SC/04 Eximstats: fixed display of large numbers with -nvr option
4148 SC/05 Eximstats: fixed merging of reports with empty tables.
4150 SC/06 Eximstats: added the -include_original_destination flag
4152 SC/07 Eximstats: removed tabs and trailing whitespace.
4154 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4156 TK/06 MBOX spool code: Add real "From " MBOX separator line
4157 so the .eml file is really in mbox format (even though
4158 most programs do not really care). Patch from Alex Miller.
4160 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4161 The latter is generated from $received_to and is only set if the
4162 message has one envelope recipient. SA can use these headers,
4163 obviously out-of-the-box. Patch from Alex Miller.
4165 PH/08 The ${def test on a variable was returning false if the variable's
4166 value was "0", contrary to what the specification has always said!
4167 The result should be true unless the variable is empty.
4169 PH/09 The syntax error of a character other than { following "${if
4170 def:variable_name" (after optional whitespace) was not being diagnosed.
4171 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4172 accidental colon was present, for example, could give incorrect results.
4174 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4175 result is used (not including appendfile, where other changes are about
4178 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4179 This involved changing a lot of size variables from int to off_t. It
4180 should work with maildirs and everything.
4182 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4183 spamd dying while we are connected to it.
4185 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4188 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4189 function for BDB 4.3.
4191 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4193 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4194 the sources for PCRE have been re-arranged and more files are now
4197 PH/15 The code I had for printing potentially long long variables in PH/11
4198 above was not the best (it lost precision). The length of off_t variables
4199 is now inspected at build time, and an appropriate printing format (%ld
4200 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4201 to be "long long int" or "long int". This is needed for the internal
4202 formatting function string_vformat().
4204 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4205 the configuration file to be ":syslog", then the script "guesses" where
4206 the logs files are, rather than using the compiled in default. In our
4207 case the guess is not the same as the compiled default, so the script
4208 suddenly stopped working when I started to use syslog. The patch checks
4209 to see if log_file_path is "". If so, it attempts to read it from exim
4210 with no configuration file to get the compiled in version, before it
4211 falls back to the previous guessing code."
4213 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4214 implementing BATV in an Exim configuration. See NewStuff for the gory
4217 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4218 Makefile that are specific to HP-UX.
4220 PH/18 If the "use_postmaster" option was set for a recipient callout together
4221 with the "random" option, the postmaster address was used as the MAIL
4222 FROM address for the random test, but not for the subsequent recipient
4223 test. It is now used for both.
4225 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4226 patch removes a few documentation additions to RFC 3028, because the
4227 latest draft now contains them. It adds the new en;ascii-case comparator
4228 and a new error check for 8bit text in MIME parts. Comparator and
4229 require names are now matched exactly. I enabled the subaddress
4230 extension, but it is not well tested yet (read: it works for me)."
4232 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4233 rework some of the code of TK/09 above to avoid the hardwired use of
4234 "%lld" and "long long". Replaced the call to snprintf() with a call to
4237 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4238 records point to non-existent hosts", "retry timeout exceeded", and
4239 "retry time not reached for any host after a long failure period".
4241 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4242 experimental DomainKeys support:
4244 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4245 (2) On an error such as an illegally used "control", the wrong name for
4246 the control was given.
4248 These problems did NOT occur unless DomainKeys support was compiled.
4250 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4252 PH/24 Added ${if match_ip condition.
4254 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4255 databases so that it will be absolutely obvious if a crash occurs in the
4256 DB library. This is a regular occurrence (often caused by mis-matched
4259 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4260 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4261 picked up on a user's system that detects such things. There doesn't seem
4262 to be a gcc warning option for this - only an attribute that has to be
4263 put on the function's prototype. It seems that in Fedora Core 4 they have
4264 set this on a number of new functions. No doubt there will be more in due
4267 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4268 specify a transport. However, if an address that was verified by such a
4269 router was the subject of a callout, Exim crashed because it tried to
4270 read the rcpt_include_affixes from the non-existent transport. Now it
4271 just assumes that the setting of that option is false. This bug was
4272 introduced by 4.51/PH/31.
4274 PH/28 Changed -d+all to exclude +memory, because that information is very
4275 rarely of interest, but it makes the output a lot bigger. People tend to
4276 do -d+all out of habit.
4278 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4279 code in os-type was giving problems when libc.so lives in lib64, like on
4282 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4283 aren't the modern standard, and it seems that some systems' include files
4284 don't always have them. Exim was already checking for some of the newer
4285 ones like T_AAAA, and defining it itself. I've added checks for all the
4286 record types that Exim uses.
4288 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4289 not automatically generating a new one, as it is supposed to. This
4290 prevented TLS from working. If the file did exist, but contained invalid
4291 data, a new version was generated, as expected. It was only the case of a
4292 non-existent file that was broken.
4294 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4295 with a change in libdomainkeys > 0.64.
4297 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4298 from DNS. If the selector record carries the flag, it now has
4299 precedence over the domain-wide flag.
4301 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4303 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4304 the use of a port name that isn't defined in /etc/services) occurred, the
4305 message was deferred as in a normal delivery, and thus remained on the
4306 spool, instead of being failed because of the mua_wrapper setting. This
4307 is now fixed, and I tidied up some of the mua_wrapper messages at the
4310 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4311 the messages in a hash of arrays rather than using individual hashes.
4312 This is a bit cleaner and results in dramatic memory savings, albeit
4313 at a slight CPU cost.
4315 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4316 as requested by Marc Sherman.
4318 SC/10 Eximstats: added histograms for user specified patterns as requested
4321 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4323 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4324 fopen() in the content-scanning modules that did not already have it.
4330 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4331 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4333 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4335 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4337 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4338 to test Sieve filters that use "vacation".
4340 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4341 that changes the way the GnuTLS parameters are stored in the cache file.
4342 The new format can be generated externally. For backward compatibility,
4343 if the data in the cache doesn't make sense, Exim assumes it has read an
4344 old-format file, and it generates new data and writes a new file. This
4345 means that you can't go back to an older release without removing the
4348 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4349 work if there are any delivery delays because "one_time" forces the
4350 parent to be marked "delivered", so its unseen clone is never tried
4351 again. For this reason, Exim now forbids the simultaneous setting of
4354 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4355 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4356 entry if you want to know the details. The fix, however, neglected to
4357 consider the case when local delivery batching is involved. The test for
4358 "previously delivered" was not happening when checking to see if an
4359 address could be batched with a previous (undelivered) one; under
4360 certain circumstances this could lead to multiple deliveries to the same
4363 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4364 in its include files, and this causes problems building Exim.
4366 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4367 header_syntax) but Exim was just ignoring anything given after a slash.
4368 In particular, this caused confusion with an attempt to use "verify =
4369 reverse_host_lookup/defer_ok". An error is now given when options are
4370 supplied for verify items that do not have them. (Maybe reverse_host_
4371 lookup should have a defer_ok option, but that's a different point.)
4373 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4374 defined by RFC 821) to 2048, because there were problems with some AUTH
4375 commands, and RFC 1869 says the size should be increased for extended
4376 SMTP commands that take arguments.
4378 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4381 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4382 "unknown" error; now it says that the functionality isn't in the binary.
4384 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4385 an address' error message when a string expansion fails (syntax or
4386 whatever). Otherwise the password may appear in the log. Following change
4387 PH/42 below, there is no longer a chance of it appearing in a bounce
4390 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4392 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4393 of its addresses was ever tried. (Bugzilla bug #2.)
4395 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4396 the result incorrectly in the debug output. (It correctly added a newline
4397 to what was transported.)
4399 TF/01 Added $received_time.
4401 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4402 commented out examples of how to interface to a virus scanner and to
4403 SpamAssassin. Also added commented examples of av_scanner and
4404 spamd_address settings.
4406 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4407 and controls are allowed in which ACLs. There were a couple of minor
4408 errors. Some of the entries in the conditions table (which is a table of
4409 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4410 negation of where the condition IS allowed.
4412 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4414 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4415 header file does not have a version number, so I've had to invent a new
4416 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4417 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4418 radiusclient), but it was contributed by a Radius user.
4420 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4421 files or directories, for OpenSSL.
4423 PH/19 When an Exim process that is running as root has to create an Exim log
4424 file, it does so in a subprocess that runs as exim:exim so as to get the
4425 ownership right at creation (otherwise, other Exim processes might see
4426 the file with the wrong ownership). There was no test for failure of this
4427 fork() call, which would lead to the process getting stuck as it waited
4428 for a non-existent subprocess. Forks do occasionally fail when resources
4429 run out. I reviewed all the other calls to fork(); they all seem to check
4432 PH/20 When checking for unexpected SMTP input at connect time (before writing
4433 the banner), Exim was not dealing correctly with a non-positive return
4434 from the read() function. If the client had disconnected by this time,
4435 the result was a log entry for a synchronization error with an empty
4436 string after "input=" when read() returned zero. If read() returned -1
4437 (an event I could not check), uninitialized data bytes were printed.
4438 There were reports of junk text (parts of files, etc) appearing after
4441 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4443 PH/22 Added support for macro redefinition, and (re)definition in between
4444 driver and ACL definitions.
4446 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4447 forgetting to use the resulting value; it was using the unexpanded value.
4449 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4450 hadn't been configured. The fix is from Juergen Kreileder, who
4451 understands it better than I do:
4453 "Here's what I see happening with three configured cyrus_sasl
4454 authenticators configured (plain, login, cram-md5):
4456 On startup auth_cyrus_sasl_init() gets called for each of these.
4457 This means three calls to sasl_listmech() without a specified mech_list.
4458 => SASL tests which mechs of all available mechs actually work
4459 => three warnings about OTP not working
4460 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4462 With the patch, sasl_listmech() also gets called three times. But now
4463 SASL's mech_list option is set to the server_mech specified in the the
4464 authenticator. Or in other words, the answer from sasl_listmech()
4465 gets limited to just the mech you're testing for (which is different
4467 => the return list contains just 'plain' or 'login', 'cram-md5' or
4468 nothing depending on the value of ob->server_mech.
4470 I've just tested the patch: Authentication still works fine,
4471 unavailable mechs specified in the exim configuration are still
4472 caught, and the auth.log warnings about OTP are gone."
4474 PH/25 When debugging is enabled, the contents of the command line are added
4475 to the debugging output, even when log_selector=+arguments is not
4478 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4479 answer is "GNU", and only if the return is "GNU/something" is the answer
4482 PH/27 $acl_verify_message is now set immediately after the failure of a
4483 verification in an ACL, and so is available in subsequent modifiers. In
4484 particular, the message can be preserved by coding like this:
4486 warn !verify = sender
4487 set acl_m0 = $acl_verify_message
4489 Previously, $acl_verify_message was set only while expanding "message"
4490 and "log_message" when a very denied access.
4492 PH/28 Modified OS/os.c-Linux with
4494 -#ifndef OS_LOAD_AVERAGE
4495 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4497 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4498 nomenclature these days.)
4500 PH/29 Installed patch from the Sieve maintainer that adds the options
4501 sieve_useraddress and sieve_subaddress to the redirect router.
4503 PH/30 In these circumstances:
4504 . Two addresses routed to the same list of hosts;
4505 . First host does not offer TLS;
4506 . First host accepts first address;
4507 . First host gives temporary error to second address;
4508 . Second host offers TLS and a TLS session is established;
4509 . Second host accepts second address.
4510 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4511 and peerdn, if requested) that were in fact used only for the second
4514 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4515 attention to any local part prefix or suffix that was matched by the
4516 router that accepted the address. It now behaves in the same way as it
4517 does for delivery: the affixes are removed from the local part unless
4518 rcpt_include_affixes is set on the transport.
4520 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4521 timeout during the DATA phase of an incoming message.
4523 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4524 applied a patch sanctioned by the Sieve maintainer.
4526 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4527 the uid or gid is negative. A case of a negative gid caused this to be
4528 noticed. The fix allows for either to be negative.
4530 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4531 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4534 PH/36 The change PH/12 above was broken. Fixed it.
4536 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4537 the grounds that routing the same address twice would always produce the
4538 same answer. This might have been true once, but it is certainly no
4539 longer true now. Routing a child address may depend on the previous
4540 routing that produced that child. Some complicated redirection strategies
4541 went wrong when messages had multiple recipients, and made Exim's
4542 behaviour dependent on the order in which the addresses were given.
4544 I have moved the duplicate checking until after the routing is complete.
4545 Exim scans the addresses that are assigned to local and remote
4546 transports, and removes any duplicates. This means that more work will be
4547 done, as duplicates will always all be routed, but duplicates are
4548 presumably rare, so I don't expect this is of any significance.
4550 For deliveries to pipes, files, and autoreplies, the duplicate checking
4551 still happens during the routing process, since they are not going to be
4554 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4555 It corrects a timeout issue with spamd. This is Ian's comment: "The
4556 background is that sometimes spamd either never reads data from a
4557 connection it has accepted, or it never writes response data. The exiscan
4558 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4559 blindly assumes that writes won't block so it may never time out."
4561 PH/39 Allow G after quota size as well as K and M.
4563 PH/40 The value set for $authenticated_id in an authenticator may not contain
4564 binary zeroes or newlines because the value is written to log lines and
4565 to spool files. There was no check on this. Now the value is run through
4566 the string_printing() function so that such characters are converted to
4567 printable escape sequences.
4569 PH/41 $message_linecount is a new variable that contains the total number of
4570 lines in the message. Compare $body_linecount, which is the count for the
4573 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4574 bounce and delay warning messages, except in certain special cases, which
4577 (a) An SMTP error message from a remote host;
4578 (b) A message specified in a :fail: redirection;
4579 (c) A message specified in a "fail" command in a system filter;
4580 (d) A message specified in a FAIL return from the queryprogram router;
4581 (e) A message specified by the cannot_route_message router option.
4583 In these cases only, Exim does include the error details in bounce and
4584 warning messages. There are also a few cases where bland messages such
4585 as "unrouteable address" or "local delivery error" are given.
4587 PH/43 $value is now also set for the "else" part of a ${run expansion.
4589 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4590 being worked on, but at least Exim now implements the latest version to
4593 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4594 process to complete was treated as a delivery failure, a timeout while
4595 writing the message to the pipe was logged, but erroneously treated as a
4596 successful delivery. Such timeouts include transport filter timeouts. For
4597 consistency with the overall process timeout, these timeouts are now
4598 treated as errors, giving rise to delivery failures by default. However,
4599 there is now a new Boolean option for the pipe transport called
4600 timeout_defer, which, if set TRUE, converts the failures into defers for
4601 both kinds of timeout. A transport filter timeout is now identified in
4604 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4605 systems where "make" and "gmake" are different, calling "gmake" at top
4606 level broke things. I've arranged for the value of $(MAKE) to be passed
4607 from the Makefile to this script so that it can call the same version of
4611 A note about Exim versions 4.44 and 4.50
4612 ----------------------------------------
4614 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4615 changes of various kinds. As a consequence, a big documentation update was
4616 needed. This delayed the release for rather longer than seemed good, especially
4617 in the light of a couple of (minor) security issues. Therefore, the changes
4618 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4619 release. So 4.44 and 4.50 are in effect two different branches that both start
4622 I have left the 4.50 change log unchanged; it contains all the changes since
4623 4.43. The change log for 4.44 is below; many of its items are identical to
4624 those for 4.50. This seems to be the most sensible way to preserve the
4625 historical information.
4631 1. Minor wording change to the doc/README.SIEVE file.
4633 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4634 computation of the current number of files was incorrect.
4636 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4637 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4640 4. Give more explanation in the error message when the command for a transport
4641 filter fails to execute.
4643 5. There are several places where Exim runs a non-Exim command in a
4644 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4645 was being done only for the command run by the queryprogram router. It is
4646 now done for all such subprocesses. The other cases are: ${run, transport
4647 filters, and the commands run by the lmtp and pipe transports.
4649 6. Added CONFIGURE_GROUP build-time option.
4651 7. Some older OS have a limit of 256 on the maximum number of file
4652 descriptors. Exim was using setrlimit() to set 1000 as a large value
4653 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4654 systems. I've change it so that if it can't get 1000, it tries for 256.
4656 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4657 was an oversight, and furthermore, ever since the addition of extra
4658 controls (e.g. 4.43/32), the checks on when to allow different forms of
4659 "control" were broken. There should now be diagnostics for all cases when a
4660 control that does not make sense is encountered.
4662 9. Added the /retain_sender option to "control=submission".
4664 10. $recipients is now available in the predata ACL (oversight).
4666 11. Tidy the search cache before the fork to do a delivery from a message
4667 received from the command line. Otherwise the child will trigger a lookup
4668 failure and thereby defer the delivery if it tries to use (for example) a
4669 cached ldap connection that the parent has called unbind on.
4671 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4672 of $address_data from the recipient verification was clobbered by the
4673 sender verification.
4675 13. The value of address_data from a sender verification is now available in
4676 $sender_address_data in subsequent conditions in the ACL statement.
4678 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4680 15. Added a new option "connect=<time>" to callout options, to set a different
4683 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4684 was its contents. (It was OK if the option was not defined at all.)
4686 17. A "Completed" log line is now written for messages that are removed from
4687 the spool by the -Mrm option.
4689 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4690 information about exactly what failed.
4692 19. Added -dd to debug only the daemon process.
4694 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4695 handles timeouts, both on the server side and network timeouts. Renamed the
4696 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4698 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4699 It is now set to "smtps".
4701 22. $host_address is now set to the target address during the checking of
4702 ignore_target_hosts.
4704 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4705 being passed; this would have caused $sender_host_name to have been used if
4706 matching the list had actually called for a host name (not very likely,
4707 since this list is usually IP addresses). A host name is now passed as
4710 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4711 code that specifies a non-restarting handler (typically sigaction() in
4712 modern systems) in an attempt to fix a rare and obscure crash bug.
4714 25. Narrowed the window for a race in the daemon that could cause it to ignore
4715 SIGCHLD signals. This is not a major problem, because they are used only to
4716 wake it up if nothing else does.
4718 26. A malformed maildirsize file could cause Exim to calculate negative values
4719 for the mailbox size or file count. Odd effects could occur as a result.
4720 The maildirsize information is now recalculated if the size or filecount
4723 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4724 support for a long time. Removed HAVE_SYS_VFS_H.
4726 28. Installed the latest version of exipick from John Jetmore.
4728 29. In an address list, if the pattern was not a regular expression, an empty
4729 subject address (from a bounce message) matched only if the pattern was an
4730 empty string. Non-empty patterns were not even tested. This was the wrong
4731 because it is perfectly reasonable to use an empty address as part of a
4732 database query. An empty address is now tested by patterns that are
4733 lookups. However, all the other forms of pattern expect the subject to
4734 contain a local part and a domain, and therefore, for them, an empty
4735 address still always fails if the pattern is not itself empty.
4737 30. Exim went into a mad DNS loop when attempting to do a callout where the
4738 host was specified on an smtp transport, and looking it up yielded more
4739 than one IP address.
4741 31. Re-factored the code for checking spool and log partition space into a
4742 function that finds that data and another that does the check. The former
4743 is then used to implement four new variables: $spool_space, $log_space,
4744 $spool_inodes, and $log_inodes.
4746 32. The RFC2047 encoding function was originally intended for short strings
4747 such as real names; it was not keeping to the 75-character limit for
4748 encoded words that the RFC imposes. It now respects the limit, and
4749 generates multiple encoded words if necessary. To be on the safe side, I
4750 have increased the buffer size for the ${rfc2047: expansion operator from
4753 33. It is now permitted to omit both strings after an "if" condition; if the
4754 condition is true, the result is "true". As before, when the second string
4755 is omitted, a false condition yields an empty string. This makes it less
4756 cumbersome to write custom ACL and router conditions.
4758 34. Failure to deliver a bounce message always caused it to be frozen, even if
4759 there was an errors_to setting on the router. The errors_to setting is now
4762 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4763 canonical form (fully expanded) before being placed in
4764 $sender_host_address.
4766 36. The table in the code that translates DNS record types into text (T_A to
4767 "A" for instance) was missing entries for NS and CNAME. It is just possible
4768 that this could have caused confusion if both these types were looked up
4769 for the same domain, because the text type is used as part of Exim's
4770 per-process caching. But the chance of anyone hitting this buglet seems
4773 37. The dnsdb lookup has been extended in a number of ways.
4775 (1) There is a new type, "zns", which walks up the domain tree until it
4776 finds some nameserver records. It should be used with care.
4778 (2) There is a new type, "mxh", which is like "mx" except that it returns
4779 just the host names, not the priorities.
4781 (3) It is now possible to give a list of domains (or IP addresses) to be
4782 looked up. The behaviour when one of the lookups defers can be
4783 controlled by a keyword.
4785 (4) It is now possible to specify the separator character for use when
4786 multiple records are returned.
4788 38. The dnslists ACL condition has been extended: it is now possible to supply
4789 a list of IP addresses and/or domains to be looked up in a particular DNS
4792 39. Added log_selector=+queue_time_overall.
4794 40. When running the queue in the test harness, wait just a tad after forking a
4795 delivery process, to get repeatability of debugging output.
4797 41. Include certificate and key file names in error message when GnuTLS fails
4798 to set them up, because the GnuTLS error message doesn't include the name
4799 of the failing file when there is a problem reading it.
4801 42. Allow both -bf and -bF in the same test run.
4803 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4805 44. The "Exiscan patch" is now merged into the mainline Exim source.
4807 45. Sometimes the final signoff response after QUIT could fail to get
4808 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4809 < 0 before doing a fflush(). This bug looks as though it goes back to the
4810 introduction of TLS in release 3.20, but "sometimes" must have been rare
4811 because the tests only now provoked it.
4813 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4814 (this can affect the format of dates).
4816 47. exim_tidydb, when checking for the continued existence of a message for
4817 which it has found a message-specific retry record, was not finding
4818 messages that were in split spool directories. Consequently, it was
4819 deleting retry records that should have stayed in existence.
4821 48. Steve fixed some bugs in eximstats.
4823 49. The SPA authentication driver was not abandoning authentication and moving
4824 on to the next authenticator when an expansion was forced to fail,
4825 contradicting the general specification for all authenticators. Instead it
4826 was generating a temporary error. It now behaves as specified.
4828 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4829 (the order specifies the preference for clients). The order is now AES256,
4830 AES128, 3DES, ARCFOUR128.
4832 51. Small patch to Sieve code - explicitly set From: when generating an
4835 52. Exim crashed if a remote delivery caused a very long error message to be
4836 recorded - for instance if somebody sent an entire SpamAssassin report back
4837 as a large number of 550 error lines. This bug was coincidentally fixed by
4838 increasing the size of one of Exim's internal buffers (big_buffer) that
4839 happened as part of the Exiscan merge. However, to be on the safe side, I
4840 have made the code more robust (and fixed the comments that describe what
4843 53. Now that there can be additional text after "Completed" in log lines (if
4844 the queue_time_overall log selector is set), a one-byte patch to exigrep
4845 was needed to allow it to recognize "Completed" as not the last thing in
4848 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4849 patch that reportedly fixes this has been added. I am not expert enough to
4850 create a test for it. This is what the patch creator wrote:
4852 "I found a little strange behaviour of ldap code when working with
4853 Windows 2003 AD Domain, where users was placed in more than one
4854 Organization Units. When I tried to give exim partial DN, the exit code
4855 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4856 But simultaneously result of request was absolutely normal ldap result,
4857 so I produce this patch..."
4859 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4860 REFERENCE, so I have modified the code to exclude the patch when that macro
4863 55. Some experimental protocols are using DNS PTR records for new purposes. The
4864 keys for these records are domain names, not reversed IP addresses. The
4865 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4866 leaves it alone. Component reversal etc. now happens only for IP addresses.
4869 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4871 57. Double the size of the debug message buffer (to 2048) so that more of very
4872 long debug lines gets shown.
4874 58. The exicyclog utility now does better if the number of log files to keep
4875 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4877 59. Two changes related to the smtp_active_hostname option:
4879 (1) $smtp_active_hostname is now available as a variable.
4880 (2) The default for smtp_banner uses $smtp_active_hostname instead
4881 of $primary_hostname.
4883 60. The host_aton() function is supposed to be passed a string that is known
4884 to be a valid IP address. However, in the case of IPv6 addresses, it was
4885 not checking this. This is a hostage to fortune. Exim now panics and dies
4886 if the condition is not met. A case was found where this could be provoked
4887 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4888 components; fortuitously, this particular loophole had already been fixed
4889 by change 4.50/55 above.
4891 If there are any other similar loopholes, the new check in host_aton()
4892 itself should stop them being exploited. The report I received stated that
4893 data on the command line could provoke the exploit when Exim was running as
4894 exim, but did not say which command line option was involved. All I could
4895 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4896 running as the user.
4899 61. There was a buffer overflow vulnerability in the SPA authentication code
4900 (which came originally from the Samba project). I have added a test to the
4901 spa_base64_to_bits() function which I hope fixes it.
4904 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4905 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4907 63. The daemon start-up calls getloadavg() while still root for those OS that
4908 need the first call to be done as root, but it missed one case: when
4909 deliver_queue_load_max is set with deliver_drop_privilege. This is
4910 necessary for the benefit of the queue runner, because there is no re-exec
4911 when deliver_drop_privilege is set.
4913 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4914 This has been fixed.
4916 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4917 was in use, was not putting the data itself into the right store pool;
4918 consequently, it could be overwritten for a subsequent message in the same
4919 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4922 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4924 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4925 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4926 the function were treating the return as a boolean value, which happened to
4927 work because 0=false and not-0=true, but is not correct code.
4929 68. The host_aton() function was not handling scoped IPv6 addresses (those
4930 with, for example, "%eth0" on the end) correctly.
4932 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4933 negated items (that is, ~something) in unsigned ints. Some compilers
4934 apparently mutter when there is no cast.
4936 70. If an address verification called from an ACL failed, and did not produce a
4937 user-specific message (i.e. there was only a "system" message), nothing was
4938 put in $acl_verify_message. In this situation, it now puts the system
4941 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4942 session; change 4.31/43 added the unwanted input to the log line - except
4943 that it did not do this in the start of session case. It now does.
4945 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4946 This is wrong and can cause the other end to generate a synchronization
4947 error if it is another Exim or anything else that does the synchronization
4948 check. A QUIT command is no longer sent after a timeout.
4950 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4951 during host lookups.
4953 74. The defer_ok option of callout verification was not working if it was used
4954 when verifying addresses in header lines, that is, for this case:
4956 verify = header_sender/callout=defer_ok
4958 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4959 those file descriptors could be used for SMTP connections. If anything
4960 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4961 could be sent to the SMTP client, causing chaos. The daemon now opens
4962 stdin, stdout, and stderr to /dev/null when it puts itself into the
4965 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4966 log by default. The user can override this with suitable Perl magic.
4968 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4969 the log message when discard triggers, was not working for the DATA ACL or
4970 for the non-SMTP ACL.
4972 78. Error message wording change in sieve.c.
4974 79. If smtp_accept_max_per_host was set, the number of connections could be
4975 restricted to fewer than expected, because the daemon was trying to set up
4976 a new connection before checking whether the processes handling previous
4977 connections had finished. The check for completed processes is now done
4978 earlier. On busy systems, this bug wouldn't be noticed because something
4979 else would have woken the daemon, and it would have reaped the completed
4982 80. If a message was submitted locally by a user whose login name contained one
4983 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4984 It caused a spool format error. I have fixed the spool reading code. A
4985 related problem was that the "from" clause in the Received: line became
4986 illegal because of the space(s). It is now covered by ${quote_local_part.
4988 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4991 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4993 83. Patch from Sieve maintainer to fix unterminated string problem in
4994 "vacation" handling.
4996 84. Some minor changes to the Linux configuration files to help with other
4997 OS variants using glibc.
4999 85. One more patch for Sieve to update vacation handling to latest spec.
5002 ----------------------------------------------------
5003 See the note above about the 4.44 and 4.50 releases.
5004 ----------------------------------------------------
5010 1. Change 4.43/35 introduced a bug that caused file counts to be
5011 incorrectly computed when quota_filecount was set in an appendfile
5014 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5015 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5018 3. Give more explanation in the error message when the command for a transport
5019 filter fails to execute.
5021 4. There are several places where Exim runs a non-Exim command in a
5022 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5023 was being done only for the command run by the queryprogram router. It is
5024 now done for all such subprocesses. The other cases are: ${run, transport
5025 filters, and the commands run by the lmtp and pipe transports.
5027 5. Some older OS have a limit of 256 on the maximum number of file
5028 descriptors. Exim was using setrlimit() to set 1000 as a large value
5029 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5030 systems. I've change it so that if it can't get 1000, it tries for 256.
5032 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5033 was an oversight, and furthermore, ever since the addition of extra
5034 controls (e.g. 4.43/32), the checks on when to allow different forms of
5035 "control" were broken. There should now be diagnostics for all cases when a
5036 control that does not make sense is encountered.
5038 7. $recipients is now available in the predata ACL (oversight).
5040 8. Tidy the search cache before the fork to do a delivery from a message
5041 received from the command line. Otherwise the child will trigger a lookup
5042 failure and thereby defer the delivery if it tries to use (for example) a
5043 cached ldap connection that the parent has called unbind on.
5045 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5046 of $address_data from the recipient verification was clobbered by the
5047 sender verification.
5049 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5050 was its contents. (It was OK if the option was not defined at all.)
5052 11. A "Completed" log line is now written for messages that are removed from
5053 the spool by the -Mrm option.
5055 12. $host_address is now set to the target address during the checking of
5056 ignore_target_hosts.
5058 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5059 being passed; this would have caused $sender_host_name to have been used if
5060 matching the list had actually called for a host name (not very likely,
5061 since this list is usually IP addresses). A host name is now passed as
5064 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5065 code that specifies a non-restarting handler (typically sigaction() in
5066 modern systems) in an attempt to fix a rare and obscure crash bug.
5068 15. Narrowed the window for a race in the daemon that could cause it to ignore
5069 SIGCHLD signals. This is not a major problem, because they are used only to
5070 wake it up if nothing else does.
5072 16. A malformed maildirsize file could cause Exim to calculate negative values
5073 for the mailbox size or file count. Odd effects could occur as a result.
5074 The maildirsize information is now recalculated if the size or filecount
5077 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5078 support for a long time. Removed HAVE_SYS_VFS_H.
5080 18. Updated exipick to current release from John Jetmore.
5082 19. Allow an empty sender to be matched against a lookup in an address list.
5083 Previously the only cases considered were a regular expression, or an
5086 20. Exim went into a mad DNS lookup loop when doing a callout where the
5087 host was specified on the transport, if the DNS lookup yielded more than
5090 21. The RFC2047 encoding function was originally intended for short strings
5091 such as real names; it was not keeping to the 75-character limit for
5092 encoded words that the RFC imposes. It now respects the limit, and
5093 generates multiple encoded words if necessary. To be on the safe side, I
5094 have increased the buffer size for the ${rfc2047: expansion operator from
5097 22. Failure to deliver a bounce message always caused it to be frozen, even if
5098 there was an errors_to setting on the router. The errors_to setting is now
5101 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5102 canonical form (fully expanded) before being placed in
5103 $sender_host_address.
5105 24. Updated eximstats to version 1.33
5107 25. Include certificate and key file names in error message when GnuTLS fails
5108 to set them up, because the GnuTLS error message doesn't include the name
5109 of the failing file when there is a problem reading it.
5111 26. Expand error message when OpenSSL has problems setting up cert/key files.
5114 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5115 (this can affect the format of dates).
5117 28. exim_tidydb, when checking for the continued existence of a message for
5118 which it has found a message-specific retry record, was not finding
5119 messages that were in split spool directories. Consequently, it was
5120 deleting retry records that should have stayed in existence.
5122 29. eximstats updated to version 1.35
5123 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5124 1.35 - bugfix such that pie charts by volume are generated correctly
5126 30. The SPA authentication driver was not abandoning authentication and moving
5127 on to the next authenticator when an expansion was forced to fail,
5128 contradicting the general specification for all authenticators. Instead it
5129 was generating a temporary error. It now behaves as specified.
5131 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5132 (the order specifies the preference for clients). The order is now AES256,
5133 AES128, 3DES, ARCFOUR128.
5135 31. Small patch to Sieve code - explicitly set From: when generating an
5138 32. Exim crashed if a remote delivery caused a very long error message to be
5139 recorded - for instance if somebody sent an entire SpamAssassin report back
5140 as a large number of 550 error lines. This bug was coincidentally fixed by
5141 increasing the size of one of Exim's internal buffers (big_buffer) that
5142 happened as part of the Exiscan merge. However, to be on the safe side, I
5143 have made the code more robust (and fixed the comments that describe what
5146 33. Some experimental protocols are using DNS PTR records for new purposes. The
5147 keys for these records are domain names, not reversed IP addresses. The
5148 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5149 leaves it alone. Component reversal etc. now happens only for IP addresses.
5152 34. The host_aton() function is supposed to be passed a string that is known
5153 to be a valid IP address. However, in the case of IPv6 addresses, it was
5154 not checking this. This is a hostage to fortune. Exim now panics and dies
5155 if the condition is not met. A case was found where this could be provoked
5156 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5157 components; fortuitously, this particular loophole had already been fixed
5158 by change 4.50/55 or 4.44/33 above.
5160 If there are any other similar loopholes, the new check in host_aton()
5161 itself should stop them being exploited. The report I received stated that
5162 data on the command line could provoke the exploit when Exim was running as
5163 exim, but did not say which command line option was involved. All I could
5164 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5165 running as the user.
5168 35. There was a buffer overflow vulnerability in the SPA authentication code
5169 (which came originally from the Samba project). I have added a test to the
5170 spa_base64_to_bits() function which I hope fixes it.
5173 36. The daemon start-up calls getloadavg() while still root for those OS that
5174 need the first call to be done as root, but it missed one case: when
5175 deliver_queue_load_max is set with deliver_drop_privilege. This is
5176 necessary for the benefit of the queue runner, because there is no re-exec
5177 when deliver_drop_privilege is set.
5179 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5180 was in use, was not putting the data itself into the right store pool;
5181 consequently, it could be overwritten for a subsequent message in the same
5182 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5185 38. Sometimes the final signoff response after QUIT could fail to get
5186 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5187 < 0 before doing a fflush(). This bug looks as though it goes back to the
5188 introduction of TLS in release 3.20, but "sometimes" must have been rare
5189 because the tests only now provoked it.
5195 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5196 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5197 Now it returns an integer. A number of calls were still expecting a T/F
5198 return. Fortuitously, in all cases, the tests worked in OK situations,
5199 which is the norm. However, things would have gone wrong on any write
5200 failures on the smtp file descriptor. This function is used when sending
5201 messages over SMTP and also when doing verify callouts.
5203 2. When Exim is called to do synchronous delivery of a locally submitted
5204 message (the -odf or -odi options), it no longer closes stderr before doing
5207 3. Implemented the mua_wrapper option.
5209 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5211 5. Implemented the functions header_remove(), header_testname(),
5212 header_add_at_position(), and receive_remove_recipient(), and exported them
5215 6. If an ACL "warn" statement specified the addition of headers, Exim already
5216 inserted X-ACL-Warn: at the start if there was no header name. However, it
5217 was not making this test for the second and subsequent header lines if
5218 there were newlines in the string. This meant that an invalid header could
5219 be inserted if Exim was badly configured.
5221 7. Allow an ACL "warn" statement to add header lines at the start or after all
5222 the Received: headers, as well as at the end.
5224 8. Added the rcpt_4xx retry error code.
5226 9. Added postmaster_mailfrom=xxx to callout verification option.
5228 10. Added mailfrom=xxxx to the callout verification option, for verify=
5231 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5232 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5234 12. Inserted some casts to stop certain compilers warning when using pointer
5235 differences as field lengths or precisions in printf-type calls (mostly
5236 affecting debugging statements).
5238 13. Added optional readline() support for -be (dynamically loaded).
5240 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5241 same clock tick as a message's arrival, so that its received time was the
5242 same as the "first fail" time on the retry record, and that message
5243 remained on the queue past the ultimate address timeout, every queue runner
5244 would try a delivery (because it was past the ultimate address timeout) but
5245 after another failure, the ultimate address timeout, which should have then
5246 bounced the address, did not kick in. This was a "< instead of <=" error;
5247 in most cases the first failure would have been in the next clock tick
5248 after the received time, and all would be well.
5250 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5251 being recognized when the domain list was tested by the match_domain
5252 condition in an expansion string.
5254 16. Added the ${str2b64: operator.
5256 17. Exim was always calling setrlimit() to set a large limit for the number of
5257 processes, without checking whether the existing limit was already
5258 adequate. (It did check for the limit on file descriptors.) Furthermore,
5259 errors from getrlimit() and setrlimit() were being ignored. Now they are
5260 logged to the main and panic logs, but Exim does carry on, to try to do its
5261 job under whatever limits there are.
5263 18. Imported PCRE 5.0.
5265 19. Trivial typo in log message " temporarily refused connection" (the leading
5268 20. If the log selector return_path_on_delivery was set and an address was
5269 redirected to /dev/null, the delivery process crashed because it assumed
5270 that a return path would always be set for a "successful" delivery. In this
5271 case, the whole delivery is bypassed as an optimization, and therefore no
5274 21. Internal re-arrangement: the function for sending a challenge and reading
5275 a response while authentication was assuming a zero-terminated challenge
5276 string. It's now changed to take a pointer and a length, to allow for
5277 binary data in such strings.
5279 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5281 23. Exim was not respecting finduser_retries when seeking the login of the
5282 uid under which it was called; it was always trying 10 times. (The default
5283 setting of finduser_retries is zero.) Also, it was sleeping after the final
5284 failure, which is pointless.
5286 24. Implemented tls_on_connect_ports.
5288 25. Implemented acl_smtp_predata.
5290 26. If the domain in control=submission is set empty, Exim assumes that the
5291 authenticated id is a complete email address when it generates From: or
5292 Sender: header lines.
5294 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5295 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5296 chown and chgrp in /bin and hostname in /usr/bin.
5298 28. Exim was keeping the "process log" file open after each use, just as it
5299 does for the main log. This opens the possibility of it remaining open for
5300 long periods when the USR1 signal hits a daemon. Occasional processlog
5301 errors were reported, that could have been caused by this. Anyway, it seems
5302 much more sensible not to leave this file open at all, so that is what now
5305 29. The long-running daemon process does not normally write to the log once it
5306 has entered its main loop, and it closes the log before doing so. This is
5307 so that log files can straightforwardly be renamed and moved. However,
5308 there are a couple of unusual error situations where the daemon does write
5309 log entries, and I had neglected to close the log afterwards.
5311 30. The text of an SMTP error response that was received during a remote
5312 delivery was being truncated at 512 bytes. This is too short for some of
5313 the long messages that one sometimes sees. I've increased the limit to
5316 31. It is now possible to make retry rules that apply only when a message has a
5317 specific sender, in particular, an empty sender.
5319 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5320 it possible to be selective about when SMTP synchronization is enforced.
5322 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5324 32. Implemented hosts_connection_nolog.
5326 33. Added an ACL for QUIT.
5328 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5331 35. Added mailbox_size and mailbox_filecount to appendfile.
5333 36. Added control = no_multiline_responses to ACLs.
5335 37. There was a bug in the logic of the code that waits for the clock to tick
5336 in the case where the clock went backwards by a substantial amount such
5337 that the microsecond fraction of "now" was more than the microsecond
5338 fraction of "then" (but the whole seconds number was less).
5340 38. Added support for the libradius Radius client library this is found on
5341 FreeBSD (previously only the radiusclient library was supported).
5347 1. When certain lookups returned multiple values in the form name=value, the
5348 quoting of the values was not always being done properly. Specifically:
5349 (a) If the value started with a double quote, but contained no whitespace,
5351 (b) If the value contained whitespace other than a space character (i.e.
5352 tabs or newlines or carriage returns) it was not quoted.
5353 This fix has been applied to the mysql and pgsql lookups by writing a
5354 separate quoting function and calling it from the lookup code. The fix
5355 should probably also be applied to nisplus, ibase and oracle lookups, but
5356 since I cannot test any of those, I have not disturbed their existing code.
5358 2. A hit in the callout cache for a specific address caused a log line with no
5359 reason for rejecting RCPT. Now it says "Previous (cached) callout
5360 verification failure".
5362 3. There was an off-by-one bug in the queryprogram router. An over-long
5363 return line was truncated at 256 instead of 255 characters, thereby
5364 overflowing its buffer with the terminating zero. As well as fixing this, I
5365 have increased the buffer size to 1024 (and made a note to document this).
5367 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5368 when Exim is waiting for an SMTP response from a remote server, Exim
5369 restarts its select() call on the socket, thereby resetting its timeout.
5370 This is not a problem when such interrupts are rare. Somebody set up a cron
5371 job to run exiwhat every 2 minutes, which is less than the normal select()
5372 timeout (5 or 10 minutes). This meant that the select() timeout never
5373 kicked in because it was always reset. I have fixed this by comparing the
5374 time when an interrupt arrives with the time at the start of the first call
5375 to select(). If more time than the timeout has elapsed, the interrupt is
5376 treated as a timeout.
5378 5. Some internal re-factoring in preparation for the addition of Sieve
5379 extensions (by MH). In particular, the "personal" test is moved to a
5380 separate function, and given an option for scanning Cc: and Bcc: (which is
5381 not set for Exim filters).
5383 6. When Exim created an email address using the login of the caller as the
5384 local part (e.g. when creating a From: or Sender: header line), it was not
5385 quoting the local part when it contained special characters such as @.
5387 7. Installed new OpenBSD configuration files.
5389 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5390 try to make them clearer.
5392 9. Callout options, other than the timeout value, were being ignored when
5393 verifying sender addresses in header lines. For example, when using
5395 verify = header_sender/callout=no_cache
5397 the cache was (incorrectly) being used.
5399 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5400 only the Cygwin environment.
5402 11. When return_path_on_delivery was set as a log selector, if different remote
5403 addresses in the same message used different return paths and parallel
5404 remote delivery occurred, the wrong values would sometimes be logged.
5405 (Whenever a remote delivery process finished, the return path value from
5406 the most recently started remote delivery process was logged.)
5408 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5409 lines when AUTH and/or TLS are in use. This is the "received protocol"
5410 field. Exim used to use "asmtp" for authenticated SMTP, without any
5411 indication (in the protocol name) for TLS use. Now it follows the RFC and
5412 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5413 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5414 names appear in log lines as well as in Received: header lines.
5416 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5417 extensions, and comparison tests, and to fix some bugs.
5419 14. Changes to the "personal" filter test:
5421 (1) The test was buggy in that it was just doing the equivalent of
5422 "contains" tests on header lines. For example, if a user's address was
5423 anne@some.where, the "personal" test would incorrectly be true for
5425 To: susanne@some.where
5427 This test is now done by extracting each address from the header in turn,
5428 and checking the entire address. Other tests that are part of "personal"
5429 are now done using regular expressions (for example, to check local parts
5430 of addresses in From: header lines).
5432 (2) The list of non-personal local parts in From: addresses has been
5433 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5434 taken from the Sieve specification recommendations.
5436 (3) If the message contains any header line starting with "List-" it is
5437 treated as non-personal.
5439 (4) The test for "circular" in the Subject: header line has been removed
5440 because it now seems ill-conceived.
5442 15. Minor typos in src/EDITME comments corrected.
5444 16. Installed latest exipick from John Jetmore.
5446 17. If headers_add on a router specified a text string that was too long for
5447 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5448 of string_sprintf() is now avoided.
5450 18. $message_body_size was not set (it was always zero) when running the DATA
5451 ACL and the local_scan() function.
5453 19. For the "mail" command in an Exim filter, no default was being set for
5454 the once_repeat time, causing a random time value to be used if "once" was
5455 specified. (If the value happened to be <= 0, no repeat happened.) The
5456 default is now 0s, meaning "never repeat". The "vacation" command was OK
5457 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5458 (I found it when inspecting the code).
5460 20. There is now an overall timeout for performing a callout verification. It
5461 defaults to 4 times the callout timeout, which applies to individual SMTP
5462 commands during the callout. The overall timeout applies when there is more
5463 than one host that can be tried. The timeout is checked before trying the
5464 next host. This prevents very long delays if there are a large number of
5465 hosts and all are timing out (e.g. when the network connections are timing
5466 out). The value of the overall timeout can be changed by specifying an
5467 additional sub-option for "callout", called "maxwait". For example:
5469 verify = sender/callout=5s,maxwait=20s
5471 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5472 to the end before writing, but this should make it even safer).
5474 22. Exim was forgetting that it had advertised PIPELINING for the second and
5475 subsequent messages on an SMTP connection. It was also not resetting its
5476 memory on STARTTLS and an internal HELO.
5478 23. When Exim logs an SMTP synchronization error within a session, it now
5479 records whether PIPELINING has been advertised or not.
5481 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5482 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5483 rather than long int.
5485 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5487 26. Added the never_mail option to autoreply.
5493 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5494 crash if the getsockname() call failed; this can happen if a connection is
5495 closed very soon after it is established. The problem was simply in the
5496 order in which certain operations were done, causing Exim to try to write
5497 to the SMTP stream before it had set up the file descriptor. The bug has
5498 been fixed by making things happen in the correct order.
5504 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5505 before the connection was closed, thus losing the rejection response.
5507 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5508 some early Solaris releases, but causes trouble in current releases where
5509 socklen_t is defined.
5511 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5514 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5517 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5518 that releasing the top of it at the end releases what was used for sub-
5519 expansions (unless the block got too big). However, discard this block if
5520 the first thing is a variable or header, so that we can use its block when
5521 it is dynamic (useful for very large $message_headers, for example).
5523 6. Lookups now cache *every* query, not just the most recent. A new, separate
5524 store pool is used for this. It can be recovered when all lookup caches are
5525 flushed. Lookups now release memory at the end of their result strings.
5526 This has involved some general refactoring of the lookup sources.
5528 7. Some code has been added to the store_xxx() functions to reduce the amount
5529 of flapping under certain conditions.
5531 8. log_incoming_interface used to affect only the <= reception log lines. Now
5532 it causes the local interface and port to be added to several more SMTP log
5533 lines, for example "SMTP connection from", and rejection lines.
5535 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5537 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5539 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5540 could be overwritten at the end of the current message (or the start of a
5541 new message if it was set in a HELO ACL). The value is now preserved for
5542 the duration of the SMTP connection.
5544 12. If a transport had a headers_rewrite setting, and a matching header line
5545 contained an unqualified address, that address was qualified, even if it
5546 did not match any rewriting rules. The underlying bug was that the values
5547 of the flags that permit the existence of unqualified sender and recipient
5548 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5549 non-local messages, and by -bnq for local messages) were not being
5550 preserved with the message after it was received.
5552 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5553 "next input=" as part of the text comprising the host identity instead of
5554 the correct text. The code was using the same buffer for two different
5555 strings. However, depending on which order the printing function evaluated
5556 its arguments, the bug did not always show up. Under Linux, for example, my
5557 test suite worked just fine.
5559 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5560 doesn't work with some older versions of Perl. It has been changed to "my",
5561 which in any case is probably the better facility to use.
5563 15. A really picky compiler found some instances of statements for creating
5564 error messages that either had too many or two few arguments for the format
5567 16. The size of the buffer for calls to the DNS resolver has been increased
5568 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5569 for addresses that have a lot of PTR records. This alleviates a problem; it
5570 does not fully solve it.
5572 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5573 buffer now truncates the list and logs the incident, which is the same
5574 action as happens when Exim is looking up a host name and its aliases.
5575 Previously in this situation something unpredictable would happen;
5576 sometimes it was "internal error: store_reset failed".
5578 18. If a server dropped the connection unexpectedly when an Exim client was
5579 using GnuTLS and trying to read a response, the client delivery process
5580 crashed while trying to generate an error log message.
5582 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5583 string, for example:
5585 warn message = H1: something\nH2: something
5587 the text was added as a single header line from Exim's point of view
5588 though it ended up OK in the delivered message. However, searching for the
5589 second and subsequent header lines using $h_h2: did not work. This has been
5590 fixed. Similarly, if a system filter added multiple headers in this way,
5591 the routers could not see them.
5593 20. Expanded the error message when iplsearch is called with an invalid key to
5594 suggest using net-iplsearch in a host list.
5596 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5597 ACLs are no longer actually imposed (and a message to that effect is
5600 22. If a "gecos" field in a passwd entry contained escaped characters, in
5601 particular, if it contained a \" sequence, Exim got it wrong when building
5602 a From: or a Sender: header from that name. A second bug also caused
5603 incorrect handling when an unquoted " was present following a character
5604 that needed quoting.
5606 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5607 was not being matched caselessly.
5609 24. Arranged for all hyphens in the exim.8 source to be escaped with
5612 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5613 properly. Recipient callout cache records were still being keyed to include
5614 the sender, even when use_sender was set false. This led to far more
5615 callouts that were necessary. The sender is no longer included in the key
5616 when use_sender is false.
5618 26. Added "control = submission" modifier to ACLs.
5620 27. Added the ${base62d: operator to decode base 62 numbers.
5622 28. dnsdb lookups can now access SRV records.
5624 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5625 the configuration file.
5627 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5628 (-v) mode. This makes the output for a verbose queue run more intelligible.
5630 31. Added a use_postmaster feature to recipient callouts.
5632 32. Added the $body_zerocount variable, containing the number of binary zero
5633 bytes in the message body.
5635 33. The time of last modification of the "new" subdirectory is now used as the
5636 "mailbox time last read" when there is a quota error for a maildir
5639 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5641 35. Added +ignore_unknown as a special item in host lists.
5643 36. Code for decoding IPv6 addresses in host lists is now included, even if
5644 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5645 address was recognized as an IP address, but was then not correctly decoded
5646 into binary, causing unexpected and incorrect effects when compared with
5653 1. Very minor rewording of debugging text in manualroute to say "list of
5654 hosts" instead of "hostlist".
5656 2. If verify=header_syntax was set, and a header line with an unqualified
5657 address (no domain) and a large number of spaces between the end of the
5658 name and the colon was received, the reception process suffered a buffer
5659 overflow, and (when I tested it) crashed. This was caused by some obsolete
5660 code that should have been removed. The fix is to remove it!
5662 3. When running in the test harness, delay a bit after writing a bounce
5663 message to get a bit more predictability in the log output.
5665 4. Added a call to search_tidyup() just before forking a reception process. In
5666 theory, someone could use a lookup in the expansion of smtp_accept_max_
5667 per_host which, without the tidyup, could leave open a database connection.
5669 5. Added the variables $recipient_data and $sender_data which get set from a
5670 lookup success in an ACL "recipients" or "senders" condition, or a router
5671 "senders" option, similar to $domain_data and $local_part_data.
5673 6. Moved the writing of debug_print from before to after the "senders" test
5676 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5677 problems for message scanning, either using a data ACL, or using
5678 local_scan() because the Received: header was not generated till after they
5679 were called (in order to set the time as the time of reception completion).
5680 I have revised the way this works. The header is now generated after the
5681 body is received, but before the ACL or local_scan() are called. After they
5682 are run, the timestamp in the header is updated.
5688 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5689 before starting a queue runner without re-exec. This happened only when
5690 deliver_drop_privilege was set or when the Exim user was set to root. The
5691 effect of the bug was that timeouts during subsequent deliveries caused
5692 crashes instead of being properly handled. The handler is now left at its
5693 default (and expected) setting.
5695 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5696 message, again when deliver_drop_privilege is set or Exim is run as root.
5697 The bug described in (1) was not present in this case, but the tidying up
5698 of the other signals was missing. I have made the two cases consistent.
5700 3. The ignore_target_hosts setting on a manualroute router was being ignored
5701 for hosts that were looked up using the /MX notation.
5703 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5706 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5707 operated on the sender address. After changing the $sender_address to <>
5708 for the sender address verify, Exim was re-instated it as the original
5709 (before rewriting) address, but remembering that it had rewritten it, so it
5710 wasn't rewriting it again. This bug also had the effect of breaking the
5711 sender address verification caching when the sender address was rewritten.
5713 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5714 This has been changed so that if the ip literal address matches
5715 ignore_target_hosts, the router declines.
5717 7. Added expansion conditions match_domain, match_address, and match_local_
5718 part (NOT match_host).
5720 8. The placeholder for the Received: header didn't have a length field set.
5722 9. Added code to Exim itself and to exim_lock to test for a specific race
5723 condition that could lead to file corruption when using MBX delivery. The
5724 issue is with the lockfile that is created in /tmp. If this file is removed
5725 after a process has opened it but before that process has acquired a lock,
5726 there is the potential for a second process to recreate the file and also
5727 acquire a lock. This could lead to two Exim processes writing to the file
5728 at the same time. The added code performs the same test as UW imapd; it
5729 checks after acquiring the lock that its file descriptor still refers to
5730 the same named file.
5732 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5733 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5736 11. Added the smtp_active_hostname option. If used, this will typically be made
5737 to depend on the incoming interface address. Because $interface_address is
5738 not set up until the daemon has forked a reception process, error responses
5739 that can happen earlier (such as "too many connections") no longer contain
5742 12. If an expansion in a condition on a "warn" statement fails because a lookup
5743 defers, the "warn" statement is abandoned, and the next ACL statement is
5744 processed. Previously this caused the whole ACL to be aborted.
5746 13. Added the iplsearch lookup type.
5748 14. Added ident_timeout as a log selector.
5750 15. Added tls_certificate_verified as a log selector.
5752 16. Added a global option tls_require_ciphers (compare the smtp transport
5753 option of the same name). This controls incoming TLS connections.
5755 17. I finally figured out how to make tls_require_ciphers do a similar thing
5756 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5757 before starting the TLS session.
5759 18. Tabs are now shown as \t in -bP output.
5761 19. If the log selector return_path_on_delivery was set, Exim crashed when
5762 bouncing a message because it had too many Received: header lines.
5764 20. If two routers both had headers_remove settings, and the first one included
5765 a superfluous trailing colon, the final name in the first list and the
5766 first name in the second list were incorrectly joined into one item (with a
5767 colon in the middle).
5773 1. Added -C and -D options to the exinext utility, mainly to make it easier
5774 to include in the automated testing, but these could be helpful when
5775 multiple configurations are in use.
5777 2. The exinext utility was not formatting the output nicely when there was
5778 an alternate port involved in the retry record key, nor when there was a
5779 message id as well (for retries that were specific to a specific message
5780 and a specific host). It was also confused by IPv6 addresses, because of
5781 the additional colons they contain. I have fixed the IPv4 problem, and
5782 patched it up to do a reasonable job for IPv6.
5784 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5785 delivery, the log line now contains "pipelined" if PIPELINING was used.
5787 4. An SMTP transport process used to panic and die if the bind() call to set
5788 an explicit outgoing interface failed. This has been changed; it is now
5789 treated in the same way as a connect() failure.
5791 5. A reference to $sender_host_name in the part of a conditional expansion
5792 that was being skipped was still causing a DNS lookup. This no longer
5795 6. The def: expansion condition was not recognizing references to header lines
5796 that used bh_ and bheader_.
5798 7. Added the _cache feature to named lists.
5800 8. The code for checking quota_filecount in the appendfile transport was
5801 allowing one more file than it should have been.
5803 9. For compatibility with Sendmail, the command line option
5811 and sets the incoming protocol and host name (for trusted callers). The
5812 host name and its colon can be omitted when only the protocol is to be set.
5813 Note the Exim already has two private options, -pd and -ps, that refer to
5814 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5815 "s", but I don't think that's a major issue.
5817 10. A number of refactoring changes to the code, none of which should affect
5820 (a) The number of logging options was getting close to filling up the
5821 32-bit word that was used as a bit map. I have split them into two classes:
5822 those that are passed in the argument to log_write(), and those that are
5823 only ever tested independently outside of that function. These are now in
5824 separate 32-bit words, so there is plenty of room for expansion again.
5825 There is no change in the user interface or the logging behaviour.
5827 (b) When building, for example, log lines, the code previously used a
5828 macro that called string_cat() twice, in order to add two strings. This is
5829 not really sufficiently general. Furthermore, there was one instance where
5830 it was actually wrong because one of the argument was used twice, and in
5831 one call a function was used. (As it happened, calling the function twice
5832 did not affect the overall behaviour.) The macro has been replaced by a
5833 function that can join an arbitrary number of extra strings onto a growing
5836 (c) The code for expansion conditions now uses a table and a binary chop
5837 instead of a serial search (which was left over from when there were very
5838 few conditions). Also, it now recognizes conditions like "pam" even when
5839 the relevant support is not compiled in: a suitably worded error message is
5840 given if an attempt is made to use such a condition.
5842 11. Added ${time_interval:xxxxx}.
5844 12. A bug was causing one of the ddress fields not to be passed back correctly
5845 from remote delivery subprocesses. The field in question was not being
5846 subsequently used, so this caused to problems in practice.
5848 13. Added new log selectors queue_time and deliver_time.
5850 14. Might have fixed a bug in maildirsizefile handling that threw up
5851 "unexpected character" debug warnings, and recalculated the data
5852 unnecessarily. In any case, I expanded the warning message to give more
5855 15. Added the message "Restricted characters in address" to the statements in
5856 the default ACL that block characters like @ and % in local parts.
5858 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5859 Three changes have been made:
5861 (a) There was a serious bug; a negative response to MAIL caused the whole
5862 recipient domain to be cached as invalid, thereby blocking all messages
5863 to all local parts at the same domain, from all senders. This bug has
5864 been fixed. The domain is no longer cached after a negative response to
5865 MAIL if the sender used is not empty.
5867 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5870 (c) A new callout option, "use_sender" has been added for people who want
5871 the modified behaviour.
5877 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5880 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5881 indeed breaks things for older releases.
5883 3. Added additional logging to the case where there is a problem reading data
5884 from a filter that is running in a subprocess using a pipe, in order to
5885 try to track down a specific problem.
5887 4. Testing facility fudge: when running in the test harness and attempting
5888 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5889 getting "No route to host". Convert this to a timeout.
5891 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5894 6. Some OS don't have socklen_t but use size_t instead. This affects the
5895 fifth argument of getsockopt() amongst other things. This is now
5896 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5897 can be set for individual OS. I have set it for SunOS5, OSF1, and
5898 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5899 some earlier ones do not.
5901 7. Change 4.30/15 was not doing the test caselessly.
5903 8. The standard form for an IPv6 address literal was being rejected by address
5904 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5905 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5906 this, as well as the form without the "IPv6" on the front (but only when
5907 address literals are enabled, of course).
5909 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5911 10. Exim crashed if a message with an empty sender address specified by -f
5912 encountered a router with an errors_to setting. This could be provoked only
5913 by a command such as
5917 where an empty string was supplied; "<>" did not hit this bug.
5919 11. Installed PCRE release 4.5.
5921 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5922 remained set. It is now erased.
5924 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5925 times from message ids (which are base 36 rather than the normal 62).
5927 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5928 were being counted as actual protocol errors, and logged if the log
5929 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5930 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5931 and DATA following a set of rejected RCPTs do not count as protocol errors.
5932 In other words, Exim assumes they were pipelined, though this may not
5933 actually be the case. Of course, in all cases the client gets an
5934 appropriate error code.
5936 15. If a lookup fails in an ACL condition, a message about the failure may
5937 be available; it is used if testing the ACL cannot continue, because most
5938 such messages specify what the cause of the deferral is. However, some
5939 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5940 that caused an old message to be retained and used if a later statement
5941 caused a defer, replacing the real cause of the deferral.
5943 16. If an IP address had so many PTR records that the DNS lookup buffer
5944 was not large enough to hold them, Exim could crash while trying to process
5945 the truncated data. It now detects and logs this case.
5947 17. Further to 4.21/58, another change has been made: if (and only if) the
5948 first line of a message (the first header line) ends with CRLF, a bare LF
5949 in a subsequent header line has a space inserted after it, so as not to
5950 terminate the header.
5952 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5953 unnecessarily, used atoi() instead of strtol(), and didn't check the
5954 termination when getting file sizes from file names by regex.
5956 19. Completely re-implemented the support for maildirsize files, in the light
5957 of a number of problems with the previous contributed implementation
5958 (4.30/29). In particular:
5960 . If the quota is zero, the maildirsize file is maintained, but no quota is
5963 . If the maildir directory does not exist, it is created before any attempt
5964 to write a maildirsize file.
5966 . The quota value in the file is just a cache; if the quota is changed in
5967 the transport, the new value overrides.
5969 . A regular expression is available for excluding directories from the
5972 20. The autoreply transport checks the characters in options that define the
5973 message's headers; it allows continued headers, but it was checking with
5974 isspace() after an embedded newline instead of explicitly looking for a
5977 21. If all the "regular" hosts to which an address was routed had passed their
5978 expiry times, and had not reached their retry times, the address was
5979 bounced, even if fallback hosts were defined. Now Exim should go on to try
5982 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5983 equivalent code in the SMTP transport. Some hosts send humungous responses
5984 to HELO/EHLO, more than 1024 it seems.
5986 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5987 gives compiler warnings in some environments. I've now done it "properly",
5990 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5991 (because of problems with the built-in one) was declared to return uschar *
5992 instead of char *, causing compiler failure.
5994 25. Fixed a file descriptor leak when processing alias/forward files.
5996 26. Fixed a minor format string issue in dbfn.c.
5998 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6000 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6001 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6004 29. When a sender address is verified, it is cached, to save repeating the test
6005 when there is more than one recipient in a message. However, when the
6006 verification involves a callout, it is possible for different callout
6007 options to be set for different recipients. It is too complicated to keep
6008 track of this in the cache, so now Exim always runs a verification when a
6009 callout is required, relying on the callout cache for the optimization.
6010 The overhead is duplication of the address routing, but this should not be
6013 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6014 to be verified with callout=postmaster, and the main callout worked but the
6015 postmaster check failed, the verification correctly failed. However, if a
6016 subsequent RCPT command asked for sender verification *without* the
6017 postmaster check, incorrect caching caused this verification also to fail,
6020 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6021 it was not caching the DNS options (qualify_single, search_parents) that
6022 were used when the lookup failed. A subsequent lookup with different
6023 options therefore always gave the same answer, though there were cases
6024 where it should not have. (Example: a "domains = !$mx_any" option on a
6025 dnslookup router: the "domains" option is always processed without any
6026 widening, but the router might have qualify_single set.) Now Exim uses the
6027 cached value only when the same options are set.
6029 32. Added John Jetmore's "exipick" utility to the distribution.
6031 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6032 than a timeout (e.g. a certificate is required, and is not provided), an
6033 Exim server now closes the connection immediately. Previously it waited for
6034 the client to close - but if the client is SSL, it seems that they each
6035 wait for each other, leading to a delay before one of them times out.
6037 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6038 maintained 0.8.x compatibility because I don't think many are using it, and
6039 it is clearly obsolete.
6041 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6044 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6045 client certificate was expired. A simple patch fixes this, though I don't
6046 understand the full logic of why the verify callback is called multiple
6049 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6050 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6051 which causes problems with some clients (such as the Certicom SSL Plus
6052 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6053 disables the coutermeasure allowing Eudora to connect."
6055 38. Exim was not checking that a write() to a log file succeeded. This could
6056 lead to Bad Things if a log got too big, in particular if it hit a file
6057 size limit. Exim now panics and dies if it cannot write to a log file, just
6058 as it does if it cannot open a log file.
6060 39. Modified OS/Makefile-Linux so that it now contains
6062 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6064 The two -D definitions ensure that Exim is compiled with large file
6065 support, which makes it possible to handle log files that are bigger than
6068 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6069 instance) a domain was checked against a named list that involved a lookup,
6070 causing $domain_data to be set, then another domain was checked against the
6071 same list, then the first domain was re-checked, the value of $domain_data
6072 after the final check could be wrong. In particular, if the second check
6073 failed, it could be set empty. This bug probably also applied to
6076 41. The strip_trailing_dot option was not being applied to the address given
6077 with the -f command-line option.
6079 42. The code for reading a message's header from the spool was incrementing
6080 $received_count, but never initializing it. This meant that the value was
6081 incorrect (doubled) while delivering a message in the same process in which
6082 it was received. In the most common configuration of Exim, this never
6083 happens - a fresh exec is done - but it can happen when
6084 deliver_drop_privilege is set.
6086 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6087 it now includes up to 150 characters of the unexpected data in the log
6090 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6091 and building data strings. The size of both of these buffers was 10 000
6092 bytes - far larger than anybody would *ever* want, thought I. Needless to
6093 say, somebody hit the limit. I have increased the maximum line length to
6094 20 000 and the maximum data length of concatenated lines to 100 000. I have
6095 also fixed two bugs, because there was no checking on these buffers. Tsk,
6096 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6097 buffer is too small.
6099 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6100 lsearch lookups. Now it does.
6102 46. When parsing a route_list item in a manualroute router, a fixed-length
6103 buffer was used for the list of hosts. I made this 1024 bytes long,
6104 thinking that nobody would ever have a list of hosts that long. Wrong.
6105 Somebody had a whole pile of complicated expansion conditions, and the
6106 string was silently truncated, leading to an expansion error. It turns out
6107 that it is easier to change to an unlimited length (owing to other changes
6108 that have happened since this code was originally written) than to build
6109 structure for giving a limitation error. The length of the item that
6110 expands into the list of hosts is now unlimited.
6112 47. The lsearch lookup could not handle data where the length of text line was
6113 more than 4095 characters. Such lines were truncated, leading to shortened
6114 data being returned. It should now handle lines of any length.
6116 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6117 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6120 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6121 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6122 no longer happens in comment lines. A list of replacements is now placed
6123 at the head of all of the source files, except those whose only change is
6124 to replace PERL_COMMAND in the very first #! line.
6126 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6127 messages on the queue, with a bottom-up merge sort, using code contributed
6128 by Michael Haardt. This should make operations like -bp somewhat faster on
6129 large queues. It won't affect queue runners, except when queue_run_in_order
6132 51. Installed eximstats 1.31 in the distribution.
6134 52. Added support for SRV lookups to the dnslookup router.
6136 53. If an ACL referred to $message_body or $message_body_end, the value was not
6137 reset for any messages that followed in the same SMTP session.
6139 54. The store-handling optimization for building very long strings was not
6140 differentiating between the different store pools. I don't think this
6141 actually made any difference in practice, but I've tidied it.
6143 55. While running the routers to verify a sender address, $sender_address
6144 was still set to the sender address. This is wrong, because when routing to
6145 send a bounce to the sender, it would be empty. Therefore, I have changed
6146 it so that, while verifying a sender address, $sender_address is set to <>.
6147 (There is no change to what happens when verifying a recipient address.)
6149 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6150 target A or AAAA records (if not already returned) without resetting the
6151 qualify_single or search_parents options of the DNS resolver. These are
6152 inappropriate in this case because the targets of MX and SRV records must
6153 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6154 target that, when qualified, matched something in the local domain. These
6155 two options are now turned off when doing these lookups.
6157 57. It seems that at least some releases of Reiserfs (which does not have the
6158 concept of a fixed number of inodes) returns zero and not -1 for the
6159 number of available inodes. This interacted badly with check_spool_inodes,
6160 which assumed that -1 was the "no such thing" setting. What I have done is
6161 to check that the total number of inodes is greater than zero before doing
6162 the test of how many are available.
6164 58. When a "warn" ACL statement has a log_message modifier, the message is
6165 remembered, and not repeated. This is to avoid a lot of repetition when a
6166 message has many recipients that cause the same warning to be written.
6167 However, Exim was preserving the list of already written lines for an
6168 entire SMTP session, which doesn't seem right. The memory is now reset if a
6169 new message is started.
6171 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6172 debugging output unless log_rewrite was also set.
6174 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6175 of (int)(handle) when we know that handle contains (void *)(-1).
6177 61. The Exim daemon panic-logs an error return when it closes the incoming
6178 connection. However "connection reset by peer" seems to be common, and
6179 isn't really an error worthy of noting specially, so that particular error
6182 62. When Exim is trying to find all the local interfaces, it used to panic and
6183 die if the ioctl to get the interface flags failed. However, it seems that
6184 on at least one OS (Solaris 9) it is possible to have an interface that is
6185 included in the list of interfaces, but for which you get a failure error
6186 for this call. This happens when the interface is not "plumbed" into a
6187 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6188 failure of the "get flags" call assumes that the interface is down.
6190 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6191 makes life easier for people who are doing arithmetic on fields extracted
6192 from dates, where you often get leading zeros that should not be
6193 interpreted as octal.
6195 64. Added qualify_domain to the redirect router, to override the global
6198 65. If a pathologically long header line contained very many addresses (the
6199 report of this problem mentioned 10 000) and each of them was rewritten,
6200 Exim could use up a very large amount of memory. (It kept on making new
6201 copies of the header line as it rewrote, and never released the old ones.)
6202 At the expense of a bit more processing, the header rewriting function has
6203 been changed so that it no longer eats memory in this way.
6205 66. The generation of the Received: header has been moved from the time that a
6206 message starts to be received, to the time that it finishes. The timestamp
6207 in the Received: header should now be very close to that of the <= log
6208 line. There are two side-effects of this change:
6210 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6211 logged header lines no longer include the local Received: line, because
6212 it has not yet been created. The same applies to a copy of the message
6213 that is returned to a non-SMTP sender when a message is rejected.
6215 (b) When a filter file is tested using -bf, no additional Received: header
6216 is added to the test message. After some thought, I decided that this
6219 This change does not affect the value of $received_for. It is still set
6220 after address rewriting, but before local_scan() is called.
6222 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6224 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6225 gave an unhelpful panic error message, and a defer error. I have managed to
6226 change this behaviour so that it now rejects any supplied certificate,
6227 which seems right, as the list of acceptable certificates is empty.
6229 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6230 gave an unhelpful defer error. I have not managed to make this reject any
6231 supplied certificates, but the error message it gives is "no certificate
6232 supplied", which is not helpful.
6234 70. exigrep's output now also includes lines that are not associated with any
6235 message, but which match the given pattern. Implemented by a patch from
6236 Martin Sluka, which also tidied up the Perl a bit.
6238 71. Recipient callout verification, like sender verification, was using <> in
6239 the MAIL FROM command. This isn't really the right thing, since the actual
6240 sender may affect whether the remote host accepts the recipient or not. I
6241 have changed it to use the actual sender in the callout; this means that
6242 the cache record is now keyed on a recipient/sender pair, not just the
6243 recipient address. There doesn't seem to be a real danger of callout loops,
6244 since a callout by the remote host to check the sender would use <>.
6245 [SEE ABOVE: changed after hitting problems.]
6247 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6248 temporary errors. However, in the case of such a code being given after
6249 the end of a data transmission (i.e. after ".") Exim was failing to write
6250 a retry record for the message. (Yes, there was some broken host that was
6251 actually sending 8xx at this point.)
6253 73. An unknown lookup type in a host list could cause Exim to panic-die when
6254 the list was checked. (An example that provoked this was putting <; in the
6255 middle of a list instead of at the start.) If this happened during a DATA
6256 ACL check, a -D file could be left lying around. This kind of configuration
6257 error no longer causes Exim to die; instead it causes a defer error. The
6258 incident is still logged to the main and panic logs.
6260 74. Buglet left over from Exim 3 conversion. The message "too many messages
6261 in one connection" was written to the rejectlog but not the mainlog, except
6262 when address rewriting (yes!) was being logged.
6264 75. Added write_rejectlog option.
6266 76. When a system filter was run not as root (that is, when system_filter_user
6267 was set), the values of the $n variables were not being returned to the
6268 main process; thus, they were not subsequently available in the $sn
6271 77. Added +return_path_on_delivery log selector.
6273 78. A connection timeout was being treated differently from recipients deferred
6274 when testing hosts_max_try with a message that was older than the host's
6275 retry timeout. (The host should not be counted, thus allowing all hosts to
6276 be tried at least once before bouncing.) This may have been the cause of an
6277 occasionally reported bug whereby a message would remain on the queue
6278 longer than the retry timeout, but would be bounced if a delivery was
6279 forced. I say "may" because I never totally pinned down the problem;
6280 setting up timeout/retry tests is difficult. See also the next item.
6282 79. The ultimate address timeout was not being applied to errors that involved
6283 a combination of host plus message (for example, a timeout on a MAIL
6284 command). When an address resolved to a number of possible hosts, and they
6285 were not all tried for each delivery (e.g. because of hosts_max_try), a
6286 message could remain on the queue longer than the retry timeout.
6288 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6291 81. Fixed an obscure SMTP outgoing bug which required at least the following
6292 conditions: (a) there was another message waiting for the same server;
6293 (b) the server returned 5xx to all RCPT commands in the first message so
6294 that the message was not completed; (c) the server dropped the connection
6295 or gave a negative response to the RSET that Exim sends to abort the
6296 transaction. The observed case was a dropped connection after DATA that had
6297 been sent in pipelining mode. That is, the server had advertised PIPELINING
6298 but was not implementing it correctly. The effect of the bug was incorrect
6299 behaviour, such as trying another host, and this could lead to a crash.
6305 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6306 and daemon.c were passed as pointers to ints; they should have been
6307 pointers to socklen_t variables (which are typically unsigned ints).
6309 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6312 3. Fixed a really odd bug that affected only the testing scheme; patching a
6313 certain fixed string in the binary changed the value of another string that
6314 happened to be identical to the end of the original first string.
6316 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6317 name", it returns that address as the IP address. On some operating
6318 systems (e.g. Solaris), it also passes back the IP address string as the
6319 "host name". However, on others (e.g. Linux), it passes back an empty
6320 string. Exim wasn't checking for this, and was changing the host name to an
6321 empty string, assuming it had been canonicalized.
6323 5. Although rare, it is permitted to have more than one PTR record for a given
6324 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6325 all the names associated with an address, because they do in Solaris.
6326 However, it seems that they do not in Linux for data that comes from the
6327 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6328 I found this out when I moved to a new Linux workstation and tried to run
6329 the Exim test suite.
6331 To get round this problem I have changed the code so that it now does its
6332 own call to the DNS to look up PTR records when searching for a host name.
6333 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6334 addresses that are only in /etc/hosts are still found.
6336 This behaviour is, however, controlled by an option called host_lookup_
6337 order, which defaults to "bydns:byaddr". If people want to use the other
6338 order, or indeed, just use one or the other means of lookup, they can
6339 specify it in this variable.
6341 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6342 some operating systems, this comes back from gethostbyaddr() as an empty
6343 string, and this is what Exim used to test for. However, it seems that in
6344 other systems, "." is yielded. Exim now tests for this case too.
6346 7. The values of check_spool_space and check_log_space are now held internally
6347 as a number of kilobytes instead of an absolute number of bytes. If a
6348 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6349 kilobyte. This means that much larger values can be stored.
6351 8. Exim monitor: an attempt to get the action menu when not actually pointing
6352 at a message produces an empty menu entitled "No message selected". This
6353 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6354 no entries in it ("Shell widget menu has zero width and/or height"). So I
6355 have added a single, blank menu entry in this case.
6357 9. Added ${quote_local_part.
6359 10. MIME decoding is now applied to the contents of Subject: header lines when
6362 11. Now that a reference to $sender_host_address automatically causes a reverse
6363 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6364 host lookup before query-style lookups in lists that might use this
6365 variable. This has therefore been abolished, and the "net-" prefix is no
6366 longer necessary for query-style lookups.
6368 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6369 have been a typo for LFLAGS, so it has been changed.
6371 13. The install script calls Exim with "-C /dev/null" in order to find the
6372 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6373 to be output. However, since Exim outputs its version number before the
6374 error, it didn't break the script. It just looked ugly. I fixed this by
6375 always allowing "-C /dev/null" if the caller is root.
6377 14. Ignore overlarge ACL variable number when reading spool file - insurance
6378 against a later release with more variables having written the file.
6380 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6381 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6382 this, as well as the form without the "IPv6" on the front.
6384 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6385 OS/Makefile-Darwin file.
6387 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6388 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6390 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6391 variable. However, some OS use ints for inodes. I've added cast to long int
6392 to get rid of the compiler warning.
6394 19. I had forgotten to lock out "/../" in configuration file names when
6395 ALT_CONFIG_PREFIX was set.
6397 20. Routers used for verification do not need to specify transports. However,
6398 if such a router generated a host list, and callout was configured, Exim
6399 crashed, because it could not find a port number from the (non-existent)
6400 transport. It now assumes port 25 in this circumstance.
6402 21. Added the -t option to exigrep.
6404 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6405 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6406 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6407 them is set, LOOKUP_LSEARCH is forced.
6409 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6411 24. Added sender and host information to the "rejected by local_scan()" log
6412 line; previously there was no indication of these.
6414 25. Added .include_if_exists.
6416 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6417 Linux. It turns out that not all file systems support this. Apparently some
6418 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6419 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6420 not supported on the file descriptor, is now ignored when Exim is trying to
6421 sync a directory. This applies only to Linux.
6423 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6425 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6426 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6429 29. Added support for maildirsize files from supplied patch (modified a bit).
6431 30. The use of :fail: followed by an empty string could lead Exim to respond to
6432 sender verification failures with (e.g.):
6434 550 Verification failed for <xxx>
6435 550 Sender verify failed
6437 where the first response line was missing the '-' that indicates it is not
6438 the final line of the response.
6440 31. The loop for finding the name of the user that called Exim had a hardwired
6441 limit of 10; it now uses the value of finduser_retries, which is used for
6442 all other user lookups.
6444 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6447 33. Exim was neglecting to zero errno before one call of strtol() when
6448 expanding a string and expecting an integer value. On some systems this
6449 resulted in spurious "integer overflow" errors. Also, it was casting the
6450 result into an int without checking.
6452 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6453 did not work. The code looks as if it has *never* worked, though it appears
6454 to have been documented since at least release 1.62. I have made it work.
6456 35. The "timeout_DNS" error in retry rules, also documented since at least
6457 1.62, also never worked. As it isn't clear exactly what this means, and
6458 clearly it isn't a major issue, I have abolished the feature by treating it
6459 as "timeout", and writing a warning to the main and panic logs.
6461 36. The display of retry rules for -brt wasn't always showing the error code
6464 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6465 timeout_connect_A, timeout_connect_MX.
6467 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6468 to the empty sender.
6470 39. The daemon was not analysing the content of -oX till after it had closed
6471 stderr and disconnected from the controlling terminal. This meant that any
6472 syntax errors were only noted on the panic log, and the return code from
6473 the command was 0. By re-arranging the code a little, I've made the
6474 decoding happen first, so such errors now appear on stderr, and the return
6475 code is 1. However, the actual setting up of the sockets still happens in
6476 the disconnected process, so errors there are still only recorded on the
6479 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6480 connections (as happens on some IP stacks) was logged at start up time as
6481 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6482 it from "IPv6 and IPv4", which means that two separate sockets are being
6485 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6486 says whether AF_INET or AF_INET6 was passed as an argument.
6488 42. Exiwhat output was messed up when time zones were included in log
6491 43. Exiwhat now gives more information about the daemon's listening ports,
6492 and whether -tls-on-connect was used.
6494 44. The "port" option of the smtp transport is now expanded.
6496 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6497 silently ignored. Now an error message is written to the main and panic
6500 46. There's a new ACL modifier called "logwrite" which writes to a log file
6501 as soon as it is encountered.
6503 47. Added $local_user_uid and $local_user_gid at routing time.
6505 48. Exim crashed when trying to verify a sender address that was being
6508 49. Exim was recognizing only a space character after ".include". It now also
6509 recognizes a tab character.
6511 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6512 extracting the relevant information from the specification. The man page no
6513 longer contains scrambled data for the -d option, and I've added a section
6514 at the front about calling Exim under different names.
6516 51. Added "extra_headers" argument to the "mail" command in filter files.
6518 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6521 53. Installed eximstats 1.29.
6523 54. Added transport_filter_timeout as a generic transport option.
6525 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6526 Cc: header lines. This was required by RFC 822, but it not required by RFC
6529 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6530 incoming messages that did not have them. Now it does so only if the
6531 message originates locally, that is, if there is no associated remote host
6532 address. When Resent- header lines are present, this applies to the Resent-
6533 lines rather than the non-Resent- lines.
6535 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6536 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6538 58. Messages for configuration errors now include the name of the main
6539 configuration file - useful now that there may be more than one file in a
6540 list (.included file names were always shown).
6542 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6543 for those rare installations that do not start the daemon as root or run it
6544 setuid root. I've cut out the call to initgroups() if the daemon is not
6547 60. The Exim user and group can now be bound into the binary as text strings
6548 that are looked up at the start of Exim's processing.
6550 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6552 62. Added $mailstore_basename variable.
6554 63. Installed patch to sieve.c from Michael Haardt.
6556 64. When Exim failed to open the panic log after failing to open the main log,
6557 the original message it was trying to log was written to stderr and debug
6558 output, but if they were not available (the usual case in production), it
6559 was lost. Now it is written to syslog before the two lines that record the
6560 failures to open the logs.
6562 65. Users' Exim filters run in subprocesses under the user's uid. It is
6563 possible for a "deliver" command or an alias in a "personal" command to
6564 provoke an address rewrite. If logging of address rewriting is configured,
6565 this fails because the process is not running as root or exim. There may be
6566 a better way of dealing with this, but for the moment (because 4.30 needs
6567 to be released), I have disabled address rewrite logging when running a
6568 filter in a non-root, non-exim process.
6574 1. The buildconfig auxiliary program wasn't quoting the value set for
6575 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6576 not defined. This bug was masked in 4.22 by the effect that was fixed in
6579 2. Some messages that were rejected after a message id was allocated were
6580 shown as "incomplete" by exigrep. It no longer does this for messages that
6581 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6583 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6584 have allow_domain_literals set, the ID did not get logged in the <= line.
6585 Domain literals are now always recognized in Message-ID: header lines.
6587 4. The first argument for a ${extract expansion item is the key name or field
6588 number. Leading and trailing spaces in this item were not being ignored,
6589 causing some misleading effects.
6591 5. When deliver_drop_privilege was set, single queue runner processes started
6592 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6593 same command in the process it spins off) were not dropping privilege.
6595 6. When the daemon running as "exim" started a queue runner, it always
6596 re-executed Exim in the spun-off process. This is a waste of effort when
6597 deliver_drop_privilege is set. The new process now just calls the
6598 queue-runner function directly.
6604 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6607 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6608 ignored. Though the use of -oP was forcing the writing of a pid file, it
6609 was always written to the default place.
6611 3. If the message "no IP address found for host xxxx" is generated during
6612 incoming verification, it is now followed by identification of the incoming
6613 connection (so you can more easily find what provoked it).
6615 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6617 5. Added some features to "harden" Exim a bit more against certain attacks:
6619 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6620 be put in Local/Makefile. This is like the never_users runtime option,
6621 but it cannot be overridden. The default setting is "root".
6623 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6624 prefix string with which any file named in a -C command line option
6627 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6628 is retained for -C and -D only if the caller of Exim is root. Without
6629 it, the exim user may also use -C and -D and retain privilege.
6631 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6632 command line option is disabled.
6634 6. Macro names set by the -D option must start with an upper case letter, just
6635 like macro names defined in the configuration file.
6637 7. Added "dereference=" facility to LDAP.
6639 8. Two instances of the typo "uknown" in the source files are fixed.
6641 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6642 the Configure-Makefile script screwed up while processing it.
6644 10. Incorporated PCRE 4.4.
6646 11. The SMTP synchronization check was not operating right at the start of an
6647 SMTP session. For example, it could not catch a HELO sent before the client
6648 waited for the greeting. There is now a check for outstanding input at the
6649 point when the greeting is written. Because of the duplex, asynchronous
6650 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6651 way, but not yet received, when the check is performed.
6653 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6654 on TCP/IP sockets, because this apparently causes some broken clients to
6657 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6658 unchanged) from the Cygwin maintainer.
6660 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6661 was supported instead of testing for mbx. Effectively a typo.
6663 15. The spa authenticator server code was not checking that the input it
6664 received was valid base64.
6666 16. The debug output line for the "set" modifier in ACLs was not showing the
6667 name of the variable that was being set.
6669 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6671 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6672 lookup on its own. Something else was needed to trigger the lookup. For
6673 example, a match in host_lookup or the need for a host name in a host list.
6674 Now, if $sender_host_name is referenced and the host name has not yet been
6675 looked up, a lookup is performed. If the lookup fails, the variable remains
6676 empty, and $host_lookup_failed is set to "1".
6678 19. Added "eqi" as a case-independent comparison operator.
6680 20. The saslauthd authentication condition could segfault if neither service
6681 nor realm was specified.
6683 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6684 error message that was logged was misleading, and incoming SMTP
6685 connections were dropped. The message is now more accurate, and temporary
6686 errors are given to SMTP connections.
6688 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6689 (except RSET) with a 421 error, until QUIT is received. However, it was
6690 failing to send a response to QUIT.
6692 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6693 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6694 ACL is now run after the helo_try_verify_hosts code.
6696 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6697 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6698 case-independent, so other case variants are also recognized). Apparently
6699 some systems use these upper case variants.
6701 25. If more than two messages were waiting for the same host, and a transport
6702 filter was specified for the transport, Exim sent two messages over the
6703 same TCP/IP connection, and then failed with "socket operation on non-
6704 socket" when it tried to send the third.
6706 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6708 27. The extern definition of crypt16() in expand.c was not being excluded when
6709 the OS had its own crypt16() function.
6711 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6712 as a preferred synonym for return_size_limit, both as an option and as an
6715 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6717 30. Changed the default configuration ACL to relax the local part checking rule
6718 for addresses that are not in any local domains. For these addresses,
6719 slashes and pipe symbols are allowed within local parts, but the sequence
6720 /../ is explicitly forbidden.
6722 31. SPA server authentication was not clearing the challenge buffer before
6725 32. log_message in a "warn" ACL statement was writing to the reject log as
6726 well as to the main log, which contradicts the documentation and doesn't
6727 seem right (because no rejection is happening). So I have stopped it.
6729 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6730 However, I am unable to do any testing of this.
6732 34. Fixed an infelicity in the appendfile transport. When checking directories
6733 for a mailbox, to see if any needed to be created, it was accidentally
6734 using path names with one or more superfluous leading slashes; tracing
6735 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6737 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6738 contents are added to the log line that is written for every discarded
6739 recipient. (Previously a log_message setting was ignored.)
6741 36. The ${quote: operator now quotes the string if it is empty.
6743 37. The install script runs exim in order to find its version number. If for
6744 some reason other than non-existence or emptiness, which it checks, it
6745 could not run './exim', it was installing it with an empty version number,
6746 i.e. as "exim-". This error state is now caught, and the installation is
6749 38. An argument was missing from the function that creates an error message
6750 when Exim fails to connect to the socket for saslauthd authentication.
6751 This could cause Exim to crash, or give a corrupted message.
6753 39. Added isip, isip4, and isip6 to ${if conditions.
6755 40. The ACL variables $acl_xx are now saved with the message, and can be
6756 accessed later in routers, transports, and filters.
6758 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6759 strings in the file are not string-expanded.
6761 42. If a MAIL command specified a SIZE value that was too large to fit into an
6762 int variable, the check against message_size_limit failed. Such values are
6763 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6764 day this will have to be increased, but I don't think I want to be around
6765 when emails are that large.
6772 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6773 iconv() is not standard in FreeBSD.
6775 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6776 IPv6 enabled. The observed symptom was a segmentation fault on return from
6777 the function os_common_find_running_interfaces() in src/os.c.
6779 3. In the check_special_case() function in daemon.c I had used "errno" as an
6780 argument name, which causes warnings on some systems. This was basically a
6781 typo, since it was named "eno" in the comments!
6783 4. The code that waits for the clock to tick (at a resolution of some fraction
6784 of a second) so as to ensure message-id uniqueness was always waiting for
6785 at least one whole tick, when it could have waited for less. [This is
6786 almost certainly not relevant at current processor speeds, where it is
6787 unlikely to ever wait at all. But we try to future-proof.]
6789 5. The function that sleeps for a time interval that includes fractions of a
6790 second contained a race. It did not block SIGALRM between setting the
6791 timer, and suspending (a couple of lines later). If the interval was short
6792 and the sigsuspend() was delayed until after it had expired, the suspension
6793 never ended. On busy systems this could lead to processes getting stuck for
6796 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6797 process, before it forks any delivery processes. The open lookup caching
6798 mechanism meant that the open file or database connection was passed into
6799 the delivery process. The problem was that delivery processes always tidy
6800 up cached lookup data. This could cause a problem for the next delivery
6801 process started by the queue runner, because the external queue runner
6802 process does not know about the closure. So the next delivery process
6803 still has data in the lookup cache. In the case of a file lookup, there was
6804 no problem because closing a file descriptor in a subprocess doesn't affect
6805 the parent. However, if the lookup was caching a connection to a database,
6806 the connection was closed, and the second delivery process was likely to
6807 see errors such as "PGSQL: query failed: server closed the connection
6808 unexpectedly". The problem has been fixed by closing all cached lookups
6809 in a queue runner before running a delivery process.
6811 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6812 seem to have the "const" qualifier which it has on other OS. I've
6815 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6816 *of the same type* (client or server) with the same public name that an
6817 error should be diagnosed.
6819 9. When Exim looked up a host name for an IP address, but failed to find the
6820 original IP address when looking up the host name (a safety check), it
6821 output the message "<ip address> does not match any IP for NULL", which was
6822 confusing, to say the least. The bug was that the host name should have
6823 appeared instead of "NULL".
6825 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6826 user that is built into the binary, and the -C or -D options is used, root
6827 privilege is dropped before the configuration file is read. In addition,
6828 logging is switched to stderr instead of the normal log files. If the
6829 configuration then re-defines the Exim user, the unprivileged environment
6830 is probably not what is expected, so Exim logs a panic warning message (but
6833 However, if deliver_drop_privilege is set, the unprivileged state may well
6834 be exactly what is intended, so the warning has been cut out in that case,
6835 and Exim is allowed to try to write to its normal log files.
6841 1. smtp_return_error_details was not giving details for temporary sender
6842 or receiver verification errors.
6844 2. Diagnose a configuration error if two authenticators have the same public
6847 3. Exim used not to create the message log file for a message until the first
6848 delivery attempt. This could be confusing when incoming messages were held
6849 for policy or load reasons. The message log file is now created at the time
6850 the message is received, and an initial "Received" line is written to it.
6852 4. The automatically generated man page for command line options had a minor
6853 bug that caused no ill effects; however, a more serious problem was that
6854 the procedure for building the man page automatically didn't always
6855 operate. Consequently, release 4.20 contains an out-of-date version. This
6856 shouldn't happen again.
6858 5. When building Exim with embedded Perl support, the script that builds the
6859 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6860 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6862 6. The freeze_tell option was not being used for messages that were frozen on
6863 arrival, either by an ACL or by local_scan().
6865 7. Added the smtp_incomplete_transaction log selector.
6867 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6868 was accepting AUTH without a new EHLO.
6870 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6871 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6874 10. Logging of TCP/IP connections (when configured) now happens in the main
6875 daemon process instead of the child process, so that the TCP/IP connection
6876 count is more accurate (but it can never be perfect).
6878 11. The use of "drop" in a nested ACL was not being handled correctly in the
6879 outer ACL. Now, if condition failure induced by the nested "drop" causes
6880 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6881 or "require"), the connection is dropped.
6883 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6884 that yield "discard" can now be used with an "accept" or a "discard" verb,
6885 but an error is generated for any others (because I can't see a useful way
6886 to define what should happen).
6888 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6889 are now processed in the same way as lines in the Exim configuration file.
6890 In particular, continuation lines are supported.
6892 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6894 15. Added -ti meaning -t -i.
6896 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6897 domains, and warn by logging if others are found.
6899 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6900 returned from GIFCONF when Exim is trying to find the list of interfaces on
6901 a host. The code in os.c has been modified to copy each ifreq to an aligned
6902 structure in all cases.
6904 Also, in some cases, the returned ifreq's were being copied to a 'struct
6905 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6906 means the last couple of bytes of an IPv6 address could be chopped if the
6907 ifreq contained only a normal sockaddr (14 bytes storage).
6909 18. Named domain lists were not supported in the hosts_treat_as_local option.
6910 An entry such as +xxxx was not recognized, and was treated as a literal
6913 19. Ensure that header lines added by a DATA ACL are included in the reject log
6914 if the ACL subsequently rejects the message.
6916 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6917 MD5 (which is deprecated).
6919 21. When testing a filter file using -bf, Exim was writing a message when it
6920 took the sender from a "From " line in the message, but it was not doing so
6921 when it took $return_path from a Return-Path: header line. It now does.
6923 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6924 with a valid header line field name (a series of printing characters
6925 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6927 23. Changed "disc" in the source to "disk" to conform to the documentation and
6928 the book and for uniformity.
6930 24. Ignore Sendmail's -Ooption=value command line item.
6932 25. When execve() failed while trying to run a command in a pipe transport,
6933 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6934 could be confused with a return value of 69 from the command itself. This
6935 has been changed to 127, the value the shell returns if it is asked to run
6936 a non-existent command. The wording for the related log line suggests a
6937 non-existent command as the problem.
6939 26. If received_header_text expands to an empty string, do not add a Received:
6940 header line to the message. (Well, it adds a token one on the spool, but
6941 marks it "old" so that it doesn't get used or transmitted.)
6943 27. Installed eximstats 1.28 (addition of -nt option).
6945 28. There was no check for failure on the call to getsockname() in the daemon
6946 code. This can fail if there is a shortage of resources on the system, with
6947 ENOMEM, for example. A temporary error is now given on failure.
6949 29. Contrary to the C standard, it seems that in some environments, the
6950 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6951 program. Exim now does this explicitly; it affects the formatting of
6952 timestamps using strftime().
6954 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6955 complaints. I've now initialized all the variables, to avoid this.
6957 32. Header lines added by a system filter were not being "seen" during
6958 transport-time rewrites.
6960 33. The info_callback() function passed to OpenSSL is set up with type void
6961 (*)(SSL *, int, int), as described somewhere. However, when calling the
6962 function (actually a macro) that sets it up, the type void(*)() is
6963 expected. I've put in a cast to prevent warnings from picky compilers.
6965 34. If a DNS black list lookup found a CNAME record, but there were no A
6966 records associated with the domain it pointed at, Exim crashed.
6968 35. If a DNS black list lookup returned more than one A record, Exim ignored
6969 all but the first. It now scans all returned addresses if a particular IP
6970 value is being sought. In this situation, the contents of the
6971 $dnslist_value variable are a list of all the addresses, separated by a
6974 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6975 to accept a host name and all its aliases if the forward lookup for any of
6976 them yielded the IP address of the incoming connection. Now it accepts only
6977 those names whose forward lookup yields the correct IP address. Any other
6978 names are discarded. This closes a loophole whereby a rogue DNS
6979 administrator could create reverse DNS records to break through a
6980 wildcarded host restriction in an ACL.
6982 37. If a user filter or a system filter that ran in a subprocess used any of
6983 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6984 the wrong values were passed to the pipe command ($thisaddress had the
6985 value of $0, $0 had the value of $1, etc). This bug was introduced by
6986 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6988 38. Improved the line breaking for long SMTP error messages from ACLs.
6989 Previously, if there was no break point between 40 and 75 characters, Exim
6990 left the rest of the message alone. Two changes have been made: (a) I've
6991 reduced the minimum length to 35 characters; (b) if it can't find a break
6992 point between 35 and 75 characters, it looks ahead and uses the first one
6993 that it finds. This may give the occasional overlong line, but at least the
6994 remaining text gets split now.
6996 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6997 file descriptors might be low, and that setting 1000 would always raise it.
6998 It turns out that in some environments, the limit is already over 1000 and
6999 that lowering it causes trouble. So now Exim takes care not to decrease it.
7001 40. When delivering a message, the value of $return_path is set to $sender_
7002 address at the start of routing (routers may change the value). By an
7003 oversight, this default was not being set up when an address was tested by
7004 -bt or -bv, which affected the outcome if any router or filter referred to
7007 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7008 log only when "message" or "log_message" are set. However, if one of the
7009 conditions was an address verification, or a call to a nested ACL, the
7010 messages generated by the underlying test were being passed through. This
7011 no longer happens. The underlying message is available in $acl_verify_
7012 message for both "message" and "log_message" expansions, so it can be
7013 passed through if needed.
7015 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7016 new expansion $bh_ to give the encoded byte string without charset
7017 translation. Translation happens only if iconv() is available; HAVE_ICONV
7018 indicates this at build time. HEADERS_CHARSET gives the charset to
7019 translate to; headers_charset can change it in the configuration, and
7020 "headers charset" can change it in an individual filter file.
7022 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7023 that creates RFC 2047 encoded "words" labels them as that charset instead
7024 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7025 expansion operator; (ii) when Exim creates a From: line for a local
7026 message; (iii) when a header line is rewritten to include a "phrase" part.
7028 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7029 buggy, causing it to skip the first lines of messages whose message ID
7030 ended in 'D'. This would not have bitten before Exim release 4.14, because
7031 message IDs were unlikely to end in 'D' before then. The effect was to have
7032 incorrect size information for certain domains.
7034 45. #include "config.h" was missing at the start of the crypt16.c module. This
7035 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7038 46. If there was a timeout during a "random" callout check, Exim treated it as
7039 a failure of the random address, and carried on sending RSET and the real
7040 address. If the delay was just some slowness somewhere, the response to the
7041 original RCPT would be taken as a response to RSET and so on, causing
7042 mayhem of various kinds.
7044 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7045 when I implemented it. It didn't allow for the fact that some option values
7046 may legitimately be negative (e.g. size_addition), and it didn't even do
7047 the right test for positive values.
7049 48. Domain names in DNS records are case-independent. Exim always looks them up
7050 in lower case. Some resolvers return domain names in exactly the case they
7051 appear in the zone file, that is, they may contain uppercase letters. Not
7052 all resolvers do this - some return always lower case. Exim was treating a
7053 change of case by a resolver as a change of domain, similar to a widening
7054 of a domain abbreviation. This triggered its re-routing code and so it was
7055 trying to route what was effectively the same domain again. This normally
7056 caused routing to fail (because the router wouldn't handle the domain
7057 twice). Now Exim checks for this case specially, and just changes the
7058 casing of the domain that it ultimately uses when it transmits the message
7061 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7064 50. If a filter generated a file delivery with a non-absolute name (possible if
7065 no home directory exists for the router), the forbid_file option was not
7068 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7069 the existing equality matching.
7071 52. Exim was using ints instead of ino_t variables in some places where it was
7072 dealing with inode numbers.
7074 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7075 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7076 called TMPDIR, and if it finds it is different, it changes its value.
7078 54. The smtp_printf() function is now made available to local_scan() so
7079 additional output lines can be written before returning. There is also an
7080 smtp_fflush() function to enable the detection of a dropped connection.
7081 The variables smtp_input and smtp_batched_input are exported to
7084 55. Changed the default runtime configuration: the message "Unknown user"
7085 has been removed from the ACL, and instead placed on the localuser router,
7086 using the cannot_route_message feature. This means that any verification
7087 failures that generate their own messages won't get overridden. Similarly,
7088 the "Unrouteable address" message that was in the ACL for unverifiable
7089 relay addresses has also been removed.
7091 56. Added hosts_avoid_esmtp to the smtp transport.
7093 57. The exicyclog script was not checking for the esoteric option
7094 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7095 will work only if exicyclog is run under the appropriate euid.
7097 58. Following a discussion on the list, the rules by which Exim recognises line
7098 endings on incoming messages have been changed. The -dropcr and drop_cr
7099 options are now no-ops, retained only for backwards compatibility. The
7100 following line terminators are recognized: LF CRLF CR. However, special
7101 processing applies to CR:
7103 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7104 nor a local message in the state where . is a terminator.
7106 (ii) If a bare CR is encountered in a header line, an extra space is added
7107 after the line terminator so as not to end the header. The reasoning
7108 behind this is that bare CRs in header lines are most likely either
7109 to be mistakes, or people trying to play silly games.
7111 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7112 was being incorrectly given as 18 bytes larger than it should have been.
7113 This is a VOB (very old bug).
7115 60. This may never have affected anything current, but just in case it has:
7116 When the local host is found other than at the start of a list of hosts,
7117 the local host, those with the same MX, and any that follow, are discarded.
7118 When the list in question was part of a longer list of hosts, the following
7119 hosts (not currently being processed) were also being discarded. This no
7120 longer happens. I'm not sure if this situation could ever has previously
7123 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7126 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7127 header. This is something that is recommended in a new Internet Draft, and
7128 is something that is documented as being done by Sendmail. There are two
7129 possible values. For messages generated by the autoreply transport, Exim
7132 Auto-Submitted: auto-replied
7134 whereas for all other generated messages (e.g. bounces) it adds
7136 Auto-Submitted: auto-generated
7138 63. The "personal" condition in filters now includes a test for the
7139 Auto-Submitted: header. If it contains the string "auto-" the message it
7140 not considered personal.
7142 64. Added rcpt_include_affixes as a generic transport option.
7144 65. Added queue_only_override (default true).
7146 66. Added the syslog_duplication option.
7148 67. If what should have been the first header line of a message consisted of
7149 a space followed by a colon, Exim was mis-interpreting it as a header line.
7150 It isn't of course - it is syntactically invalid and should therefore be
7151 treated as the start of the message body. The misbehaviour could have
7152 caused a number of strange effects, including loss of data in subsequent
7153 header lines, and spool format errors.
7155 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7156 client host had authenticated. This control can now be exercised by an ACL
7157 for more flexibility.
7159 69. By default, callouts do not happen when testing with -bh. There is now a
7160 variant, -bhc, which does actually run the callout code, including
7161 consulting and updating the callout cache.
7163 70. Added support for saslauthd authentication, courtesy of Alexander
7166 71. If statvfs() failed on the spool or log directories while checking their
7167 size for availability, Exim confusingly gave the error "space shortage".
7168 Furthermore, in debugging mode it crashed with a floating point exception.
7169 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7170 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7171 serious problem, Exim now writes to the main and panic logs when this
7172 happens, with details of the failure. It then refuses to accept the
7173 incoming message, giving the message "spool directory problem" or "log
7174 directory problem" with a 421 code for SMTP messages.
7176 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7177 0, 1, and 2 exist, because some OS complain for execs without them (see
7178 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7179 descriptors. However, the code omitted to check that the open succeeded,
7180 causing mysterious errors if for some reason the permissions on /dev/null
7181 got screwed. Now Exim writes a message to the main and panic logs, and
7182 bombs out if it can't open /dev/null.
7184 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7185 interact so that it is all more flexible. It is supposed to remain
7186 backwards compatible. Also added extra_local_interfaces.
7188 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7189 to bomb out with an assertion failure - to the client this appears as a
7190 connection drop. This problem occurs in the part of the code that was taken
7191 from the Samba project. Fortunately, the assertion is in a very simple
7192 function, so I have fixed this by reproducing the function inline in the
7193 one place where it is called, and arranging for authentication to fail
7194 instead of killing the process with assert().
7196 75. The SPA client code was not working when the server requested OEM rather
7197 than Unicode encoding.
7199 76. Added code to make require_files with a specific uid setting more usable in
7200 the case where statting the file as root fails - usually a non-root-mounted
7201 NFS file system. When this happens and the failure is EACCES, Exim now
7202 forks a subprocess and does the per-uid checking as the relevant uid.
7204 77. Added process_log_path.
7206 78. If log_file_path was not explicitly set, a setting of check_log_space or
7207 check_log_inodes was ignored.
7209 79. If a space check for the spool or log partitions fails, the incident is now
7210 logged. Of course, in the latter case the data may get lost...
7212 80. Added the %p formatting code to string_format() so that it can be used to
7213 print addresses in debug_print(). Adjusted all the address printing in the
7214 debugging in store.c to use %p rather than %d.
7216 81. There was a concern that a line of code in smtp_in.c could overflow a
7217 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7218 initially expressed, the concern was not well-founded, because trailing
7219 spaces are removed early. However, if the trailing spaces were followed by
7220 a NULL, they did not get removed, so the overflow was possible. Two fixes
7223 (a) I re-wrote the offending code in a cleaner fashion.
7224 (b) If an incoming SMTP command contains a NULL character, it is rejected
7227 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7228 runs initgroups(), so that if the Exim user is in any additional groups,
7229 they will be used during message reception.
7235 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob