1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
56 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
57 When redis_servers is set to a list of > 1 element, and the Redis servers
58 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
59 case of MOVED into a DEFER case instead, thus moving the query onto the
60 next server in the list. For a cluster of N elements, all N servers must
61 be defined in redis_servers.
63 GF/02 Catch and remove uninitialized value warning in exiqsumm
64 Check for existence of @ARGV before looking at $ARGV[0]
66 JH/01 Replace the store_release() internal interface with store_newblock(),
67 which internalises the check required to safely use the old one, plus
68 the allocate and data copy operations duplicated in both (!) of the
71 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
72 modifier. This matches the restriction on the commandline.
74 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
75 Previously only the last row was returned.
77 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
78 we assumed that tags in the header were well-formed, and parsed the
79 element content after inspecting only the first char of the tag.
80 Assumptions at that stage could crash the receive process on malformed
83 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
84 While running the DKIM ACL we operate on the Permanent memory pool so that
85 variables created with "set" persist to the DATA ACL. Also (at any time)
86 DNS lookups that fail create cache records using the Permanent pool. But
87 expansions release any allocations made on the current pool - so a dnsdb
88 lookup expansion done in the DKIM ACL releases the memory used for the
89 DNS negative-cache, and bad things result. Solution is to switch to the
90 Main pool for expansions.
91 While we're in that code, add checks on the DNS cache during store_reset,
92 active in the testsuite.
93 Problem spotted, and debugging aided, by Wolfgang Breyha.
95 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
96 When none of the hosts presented to a transport match an already-open
97 connection, close it and proceed with the list. Previously we would
98 queue the message. Spotted by Lena with Yahoo, probably involving
101 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
102 Previously a spurious "250 OK id=" response was appended to the proper
105 JH/08 The "support for" informational output now, which built with Content
106 Scanning support, has a line for the malware scanner interfaces compiled
107 in. Interface can be individually included or not at build time.
109 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
110 by the template makefile "src/EDITME". The "STREAM" support for an older
111 ClamAV interface method is removed.
113 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
114 rows affected is given instead).
116 JH/11 The runtime Berkeley DB library version is now additionally output by
117 "exim -d -bV". Previously only the compile-time version was shown.
119 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
120 SMTP connection. Previously, when one had more receipients than the
121 first, an abortive onward connection was made. Move to full support for
122 multiple onward connections in sequence, handling cutthrough connection
123 for all multi-message initiating connections.
125 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
126 routers. Previously, a multi-recipient message would fail to match the
127 onward-connection opened for the first recipient, and cause its closure.
129 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
130 a timeout on read on a GnuTLS initiating connection, resulting in the
131 initiating connection being dropped. This mattered most when the callout
132 was marked defer_ok. Fix to keep the two timeout-detection methods
135 JH/15 Relax results from ACL control request to enable cutthrough, in
136 unsupported situations, from error to silently (except under debug)
137 ignoring. This covers use with PRDR, frozen messages, queue-only and
140 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
142 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
143 metadata, resulting in a crash in free().
145 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
146 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
147 Broken also in d185889f4, with init system revamp.
149 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
150 Previously we abruptly closed the connection after reading a malware-
151 found indication; now we go on to read the "scan ok" response line,
154 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
155 ACL. Previously, a crash would result.
157 JH/19 Speed up macro lookups during configuration file read, by skipping non-
158 macro text after a replacement (previously it was only once per line) and
159 by skipping builtin macros when searching for an uppercase lead character.
161 JH/20 DANE support moved from Experimental to mainline. The Makefile control
162 for the build is renamed.
164 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
165 was allocated for every new TLS startup, meaning one per message. Fix
166 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
168 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
169 reported the original. Fix to report (as far as possible) the ACL
170 result replacing the original.
172 JH/23 Fix memory leak during multi-message connections using STARTTLS under
173 OpenSSL. Certificate information is loaded for every new TLS startup,
174 and the resources needed to be freed.
176 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
178 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
181 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
182 DATA response info to the (existing) per-recipient response info for
183 the "C=" log element. It can have useful tracking info from the
184 destination system. Patch from Simon Arlott.
186 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
187 length value. Previously this would segfault.
189 HS/02 Support Avast multiline protoocol, this allows passing flags to
190 newer versions of the scanner.
192 JH/28 Ensure that variables possibly set during message acceptance are marked
193 dead before release of memory in the daemon loop. This stops complaints
194 about them when the debug_store option is enabled. Discovered specifically
195 for sender_rate_period, but applies to a whole set of variables.
196 Do the same for the queue-runner and queue-list loops, for variables set
197 from spool message files. Do the same for the SMTP per-message loop, for
198 certain variables indirectly set in ACL operations.
200 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
201 as a multi-recipient message from a mailinglist manager). The coding had
202 an arbitrary cutoff number of characters while checking for more input;
203 enforced by writing a NUL into the buffer. This corrupted long / fast
204 input. The problem was exposed more widely when more pipelineing of SMTP
205 responses was introduced, and one Exim system was feeding another.
206 The symptom is log complaints of SMTP syntax error (NUL chars) on the
207 receiving system, and refused recipients seen by the sending system
208 (propating to people being dropped from mailing lists).
209 Discovered and pinpointed by David Carter.
211 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
212 replaced by the ${authresults } expansion.
214 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
216 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
217 allows proper process termination in container environments.
219 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
220 Previously the "final dot" had a newline after it; ensure it is CR,LF.
222 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
223 and "err_perm", deprecated since 4.83 when the RFC-defined words
224 "temperror" and "permerror" were introduced.
226 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
227 transport-filters or DKIM-signing. The restriction was lost in the
228 consolidation of verify-callout and delivery SMTP handling.
229 Extend the restriction to also cover ARC-signing.
231 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
232 in defer=pass mode supply a 450 to the initiator. Previously the message
235 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
236 tls_require_ciphers is used as before.
238 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
239 "pass_unscanned". Only tmpfails from the scanner are written to
240 the paniclog, as they may require admin intervention (permission
241 denied, license issues). Other scanner errors (like decompression
242 bombs) do not cause a paniclog entry.
244 JH/36 Fix reinitialisation of DKIM logging variable between messages.
245 Previously it was possible to log spurious information in receive log
248 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
249 triggered odd behaviour from Outlook Express clients.
251 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
254 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
255 since the IETF WG has not yet settled on that versus the original
256 "bare" representation.
258 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
259 Previously the millisecond value corrupted the output.
260 Fix also for syslog_pid=no and log_selector +pid, for which the pid
261 corrupted the output.
267 JH/01 Rework error string handling in TLS interface so that the caller in
268 more cases is responsible for logging. This permits library-sourced
269 string to be attached to addresses during delivery, and collapses
270 pairs of long lines into single ones.
272 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
273 during configuration. Wildcards are allowed and expanded.
275 JH/02 Rework error string handling in DKIM to pass more info back to callers.
276 This permits better logging.
278 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
279 do not close it down and have the child transport start it up again on
280 the passed-on TCP connection. Instead, proxy the child (and any
281 subsequent ones) for TLS via a unix-domain socket channel. Logging is
282 affected: the continued delivery log lines do not have any DNSSEC, TLS
283 Certificate or OCSP information. TLS cipher information is still logged.
285 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
286 identical IP addresses on different listening ports. Will also affect
289 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
290 add noisy ifdef guards to special-case this sillyness.
291 Patch from Bernd Kuhls.
293 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
294 than 255 are no longer allowed.
296 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
297 Disable the session-cache too, which might reduce our load. Since we
298 currrectly use a new context for every connection, both as server and
299 client, there is no benefit for these.
300 GnuTLS appears to not support tickets server-side by default (we don't
301 call gnutls_session_ticket_enable_server()) but client side is enabled
302 by default on recent versions (3.1.3 +) unless the PFS priority string
305 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
306 <https://reproducible-builds.org/specs/source-date-epoch/>.
308 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
309 the check for any unsuccessful recipients did not notice the limit, and
310 erroneously found still-pending ones.
312 JH/08 Pipeline CHUNKING command and data together, on kernels that support
313 MSG_MORE. Only in-clear (not on TLS connections).
315 JH/09 Avoid using a temporary file during transport using dkim. Unless a
316 transport-filter is involved we can buffer the headers in memory for
317 creating the signature, and read the spool data file once for the
318 signature and again for transmission.
320 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
321 4.77 as the kernel support then wasn't solid, having issues in 64bit
322 mode. Now, it's been long enough. Add support for FreeBSD also.
324 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
325 case where the routing stage had gathered several addresses to send to
326 a host before calling the transport for the first, we previously failed
327 to close down TLS in the old transport process before passing the TCP
328 connection to the new process. The new one sent a STARTTLS command
329 which naturally failed, giving a failed delivery and bloating the retry
330 database. Investigation and fix prototype from Wolfgang Breyha.
332 JH/12 Fix check on SMTP command input synchronisation. Previously there were
333 false-negatives in the check that the sender had not preempted a response
334 or prompt from Exim (running as a server), due to that code's lack of
335 awareness of the SMTP input buffering.
337 PP/04 Add commandline_checks_require_admin option.
338 Exim drops privileges sanely, various checks such as -be aren't a
339 security problem, as long as you trust local users with access to their
340 own account. When invoked by services which pass untrusted data to
341 Exim, this might be an issue. Set this option in main configuration
342 AND make fixes to the calling application, such as using `--` to stop
345 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
346 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
347 responses to those, into a single TLS record each way (this usually means
348 a single packet). As a side issue, smtp_enforce_sync now works on TLS
351 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
352 affects you only if you're dancing at the edge of the param size limits.
353 If you are, and this message makes sense to you, then: raise the
354 configured limit or use OpenSSL 1.1. Nothing we can do for older
357 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
358 cmdline element to get the documented default one. Previously it was
361 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
362 are unsuable for non-message-specific reasons - eg. connection timeout,
363 banner-time rejection.
365 JH/16 Fix logging of delivery remote port, when specified by router, under
368 PP/06 Repair manualroute's ability to take options in any order, even if one
369 is the name of a transport.
372 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
374 JH/17 Change the list-building routines interface to use the expanding-string
375 triplet model, for better allocation and copying behaviour.
377 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
378 Previously it was constructed the first time a possibly-matching string
379 was met in the configuration file input during startup; now it is done
382 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
383 compatible one, to avoid the (poorly documented) possibility of a config
384 file in the working directory redirecting the DB files, possibly correpting
385 some existing file. CVE-2017-10140 assigned for BDB.
387 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
388 cache-hot. Previously, although the result was properly cached, the
389 initial verify call returned a defer.
391 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
392 the main verify for receipient in uncached-mode.
394 JH/22 Retire historical build files to an "unsupported" subdir. These are
395 defined as "ones for which we have no current evidence of testing".
397 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
398 if present. Previously it was ignored.
400 JH/24 Start using specified-initialisers in C structure init coding. This is
401 a C99 feature (it's 2017, so now considered safe).
403 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
404 if was a fixed-sized field and bitmask ops via macros; it is now more
407 PP/07 GitHub PR 56: Apply MariaDB build fix.
408 Patch provided by Jaroslav Škarvada.
410 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
411 during Coverity cleanups [4.87 JH/47]
412 Diagnosis and fix provided by Michael Fischer v. Mollard.
414 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
415 the right size to place the terminating semicolon on its own folded
416 line, the header hash was calculated to an incorrect value thanks to
417 the (relaxed) space the fold became.
419 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
420 and confused the parent.
422 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
423 which could crash as a result. This could lead to undeliverable messages.
425 JH/28 Logging: "next input sent too soon" now shows where input was truncated
428 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
429 matters on fast-turnover and PID-randomising systems, which were getting
430 out-of-order delivery.
432 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
433 a possibly-overlapping copy. The symptom was that "Remote host closed
434 connection in response to HELO" was logged instead of the actual 4xx
437 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
438 Previously only that bufferd was discarded, resulting in SYMTP command
441 JH/32 DKIM: when a message has multiple signatures matching an identity given
442 in dkim_verify_signers, run the dkim acl once for each. Previously only
443 one run was done. Bug 2189.
445 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
446 "panic the current process" to "deliberately defer". The panic log is
447 still written with the problem list name; the mail and reject logs now
448 get a temp-reject line for the message that was being handled, saying
449 something like "domains check lookup or other defer". The SMTP 451
450 message is still "Temporary local problem".
452 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
453 A crafted sequence of BDAT commands could result in in-use memory beeing
454 freed. CVE-2017-16943.
456 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
457 from SMTP input. Previously it was always done; now only done for DATA
458 and not BDAT commands. CVE-2017-16944.
460 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
461 to the message (such as an overlong header line). Previously this was
462 not done and we did not exit BDAT mode. Followon from the previous item
463 though a different problem.
469 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
470 than -2003 did; needs libidn2 in addition to libidn.
472 JH/02 The path option on a pipe transport is now expanded before use.
474 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
475 Patch provided by "Björn", documentation fix added too.
477 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
478 missing a wire-to-host endian conversion.
480 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
481 close after a BDAT command line could be taken as a following command,
482 giving a synch failure. Fix by only checking for synch immediately
483 before acknowledging the chunk.
485 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
486 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
487 macro. Patches provided by Josh Soref.
489 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
490 Previously we did not; the RFC seems ambiguous and VRFY is not listed
491 by IANA as a service extension. However, John Klensin suggests that we
494 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
495 the dkim code may be unix-mode line endings rather than smtp wire-format
496 CRLF, so prepend a CR to any bare LF.
498 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
499 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
501 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
502 the first were themselves being wrongly included in the feed into dkim
503 processing; with most chunk sizes in use this resulted in an incorrect
504 body hash calculated value.
506 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
507 DKIM signature block, for verification. Although advised against by
508 standards it is specifically not ruled illegal.
510 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
512 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
513 missing a body hash (the bh= tag).
515 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
516 It seems that HAProxy sends the Proxy Protocol information in clear and
517 only then does a TLS startup, so do the same.
519 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
520 TCP connections (such as for Spamd) unless the daemon successfully set
521 Fast Open mode on its listening sockets. This fixes breakage seen on
522 too-old kernels or those not configured for Fast Open, at the cost of
523 requiring both directions being enabled for TFO, and TFO never being used
524 by non-daemon-related Exim processes.
526 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
527 endings, at least on the first header line. Try to canonify any that get
528 past that check, despite the cost.
530 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
531 now limited to an arbitrary five deep, while parsing addresses with the
532 strip_excess_angle_brackets option enabled.
534 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
535 instead leave the unprompted TLS handshake in socket buffer for the
536 TLS library to consume.
538 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
540 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
542 JH/16 Drop variables when they go out of scope. Memory management drops a whole
543 region in one operation, for speed, and this leaves assigned pointers
544 dangling. Add checks run only under the testsuite which checks all
545 variables at a store-reset and panics on a dangling pointer; add code
546 explicitly nulling out all the variables discovered. Fixes one known
547 bug: a transport crash, where a dangling pointer for $sending_ip_address
548 originally assigned in a verify callout, is re-used.
550 PP/06 Drop '.' from @INC in various Perl scripts.
552 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
554 PP/08 Reduce a number of compilation warnings under clang; building with
555 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
556 should be warning-free.
558 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
560 HS/01 Fix portability problems introduced by PP/08 for platforms where
561 realloc(NULL) is not equivalent to malloc() [SunOS et al].
563 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
564 chunk. This allows us to accept broken chunked messages. We need a more
565 general solution here.
567 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
568 already-broken messages in the queue.
570 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
572 JH/19 Fix reference counting bug in routing-generated-address tracking.
578 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
579 supports it and a size is available (ie. the sending peer gave us one).
581 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
582 years of being deprecated). The replacements are the ACLs
583 acl_smtp_mime and acl_not_smtp_mime.
585 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
586 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
587 or even an in-clear connection were permitted. Now, if the host lookup
588 was dnssec and dane was requested then the host is only used if the
589 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
590 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
591 if one fails this test.
592 This means that a poorly-configured remote DNS will make it incommunicado;
593 but it protects against a DNS-interception attack on it.
595 JH/04 Bug 1810: make continued-use of an open smtp transport connection
596 non-noisy when a race steals the message being considered.
598 JH/05 If main configuration option tls_certificate is unset, generate a
599 self-signed certificate for inbound TLS connections.
601 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
602 in rewrites and routers.
604 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
605 and logged a warning sing 4.83; now they are a configuration file error.
607 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
608 (lacking @domain). Apply the same qualification processing as RCPT.
610 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
612 JH/10 Support ${sha256:} applied to a string (as well as the previous
615 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
616 a cutthrough deliver is pending, as we always want to make a connection.
617 This also avoids re-routing the message when later placing the cutthrough
618 connection after a verify cache hit.
619 Do not update it with the verify result either.
621 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
622 when routing results in more than one destination address.
624 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
625 signing (which inhibits the cutthrough capability). Previously only
626 the presence of an option was tested; now an expansion evaluating as
627 empty is permissible (obviously it should depend only on data available
628 when the cutthrough connection is made).
630 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
631 the relevant preceding SMTP command did not note the pipelining mode.
633 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
634 Previously they were not counted.
636 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
637 as one having no matching records. Previously we deferred the message
638 that needed the lookup.
640 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
641 distinguished as "(=".
643 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
644 for missing MX records. Previously it only worked for missing A records.
646 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
648 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
649 after the data-go-ahead and data-ack. Patch from Jason Betts.
651 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
652 even for a "none" policy. Patch from Tony Meyer.
654 JH/22 Fix continued use of a connection for further deliveries. If a port was
655 specified by a router, it must also match for the delivery to be
658 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
659 When one of the recipients of a message was unsuitable for the connection
660 (has no matching addresses), we lost track of needing to mark it
661 deferred. As a result mail would be lost.
663 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
665 JH/25 Decoding ACL controls is now done using a binary search; the source code
666 takes up less space and should be simpler to maintain. Merge the ACL
667 condition decode tables also, with similar effect.
669 JH/26 Fix problem with one_time used on a redirect router which returned the
670 parent address unchanged. A retry would see the parent address marked as
671 delivered, so not attempt the (identical) child. As a result mail would
674 JH/27 Fix a possible security hole, wherein a process operating with the Exim
675 UID can gain a root shell. Credit to http://www.halfdog.net/ for
676 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
679 JH/28 Enable {spool,log} filesystem space and inode checks as default.
680 Main config options check_{log,spool}_{inodes,space} are now
681 100 inodes, 10MB unless set otherwise in the configuration.
683 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
684 Previously it only applied to the main-section connection policy
687 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
689 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
690 by me. Added RFC7919 DH primes as an alternative.
692 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
693 are not in the system include path.
695 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
696 GnuTLS, when a session startup failed (eg because the client disconnected)
697 Exim did stdio operations after fclose. This was exposed by a recent
698 change which nulled out the file handle after the fclose.
700 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
701 signed directly by the cert-signing cert, rather than an intermediate
702 OCSP-signing cert. This is the model used by LetsEncrypt.
704 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
706 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
707 an incoming connection.
709 HS/02 Bug 1802: Do not half-close the connection after sending a request
712 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
713 fallback to "prime256v1".
715 JH/34 SECURITY: Use proper copy of DATA command in error message.
716 Could leak key material. Remotely exploitable. CVE-2016-9963.
722 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
723 and 3.4.4 - once the server is enabled to respond to an OCSP request
724 it does even when not requested, resulting in a stapling non-aware
725 client dropping the TLS connection.
727 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
728 support variable-length bit vectors. No functional change.
730 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
731 The I= interface field on outgoing lines is now after the H= remote
732 host field, same as incoming lines. There is a separate
733 outgoing_interface log selector which allows you to disable the
736 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
737 If not running log_selector +smtp_connection the mainlog would be held
738 open indefinitely after a "too many connections" event, including to a
739 deleted file after a log rotate. Leave the per net connection logging
740 leaving it open for efficiency as that will be quickly detected by the
741 check on the next write.
743 HS/01 Bug 1671: Fix post transport crash.
744 Processing the wait-<transport> messages could crash the delivery
745 process if the message IDs didn't exist for some reason. When
746 using 'split_spool_directory=yes' the construction of the spool
747 file name failed already, exposing the same netto behaviour.
749 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
750 mime_regex ACL conditions.
752 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
753 to DSN fail messages (bounces): remote IP, remote greeting, remote response
754 to HELO, local diagnostic string.
756 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
757 log line to debug. Even when configured with a tls authenticator many
758 client connections are expected to not authenticate in this way, so
759 an authenticate fail is not an error.
761 HS/02 Add the Exim version string to the process info. This way exiwhat
762 gives some more detail about the running daemon.
764 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
765 matter for fast-change records such as DNSBLs.
767 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
768 retry record, even if constant. There may be multiple transports with
769 different interface settings and the retry behaviour needs to be kept
772 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
774 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
776 JH/10 Bug 840: fix log_defer_output option of pipe transport
778 JH/11 Bug 830: use same host for all RCPTS of a message, even under
779 hosts_randomize. This matters a lot when combined with mua_wrapper.
781 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
782 ${quote_pgsql:<string>} operator.
784 JH/13 Bug 1708: avoid misaligned access in cached lookup.
786 JH/14 Change header file name for freeradius-client. Relevant if compiling
787 with Radius support; from the Gentoo tree and checked under Fedora.
789 JH/15 Bug 1712: Introduce $prdr_requested flag variable
791 JH/16 Bug 1714: Permit an empty string as expansion result for transport
792 option transport_filter, meaning no filtering.
794 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
796 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
797 defaults to "*" (all hosts). The variable is now available when not built
798 with TLS, default unset, mainly to enable keeping the testsuite sane.
799 If a server certificate is not supplied (via tls_certificate) an error is
800 logged, and clients will find TLS connections fail on startup. Presumably
801 they will retry in-clear.
802 Packagers of Exim are strongly encouraged to create a server certificate
803 at installation time.
805 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
806 with the $config_file variable.
808 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
809 in transport context, after the attempt, and per-recipient. The latter type
810 is per host attempted. The event data is the error message, and the errno
811 information encodes the lookup type (A vs. MX) used for the (first) host,
812 and the trailing two digits of the smtp 4xx response.
814 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
815 to write to mainlog (or rejectlog, paniclog) in the window between file
816 creation and permissions/ownership being changed. Particularly affects
817 installations where exicyclog is run as root, rather than exim user;
818 result is that the running daemon panics and dies.
820 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
822 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
823 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
824 "pri" and "weight". Note that the previous implicit priority given by the
825 list order is no longer honoured.
827 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
830 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
831 by defining SUPPORT_SOCKS.
833 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
834 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
835 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
836 variables are renamed to proxy_{local,external}_{address,port}.
838 JH/27 Move Internationalisation support from Experimental to mainline, enabled
839 for a build by defining SUPPORT_I18N
841 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
842 of the query string, and make ${quote_redis:} do that quoting.
844 JH/29 Move Events support from Experimental to mainline, enabled by default
845 and removable for a build by defining DISABLE_EVENT.
847 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
849 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
850 cached by the daemon.
852 JH/32 Move Redis support from Experimental to mainline, enabled for a build
853 by defining LOOKUP_REDIS. The libhiredis library is required.
855 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
856 keys are given for lookup.
858 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
859 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
860 only supported when built with TLS support. The PolarSSL SHA routines
861 are still used when the TLS library is too old for convenient support.
863 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
864 openssl_options), for security. OpenSSL forces this from version 1.1.0
865 server-side so match that on older versions.
867 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
868 allocation for $value could be released as the expansion processing
869 concluded, but leaving the global pointer active for it.
871 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
872 and to use the domains and local_parts ACL conditions.
874 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
875 incorrectly not doubled on cutthrough transmission, hence seen as a
876 body-termination at the receiving system - resulting in truncated mails.
877 Commonly the sender saw a TCP-level error, and retransmitted the message
878 via the normal store-and-forward channel. This could result in duplicates
879 received - but deduplicating mailstores were liable to retain only the
880 initial truncated version.
882 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
884 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
886 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
887 we're in there, support oversigning also; bug 1309.
889 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
891 HS/04 Add support for keep_environment and add_environment options.
893 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
894 either intentional arithmetic overflow during PRNG, or testing config-
897 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
898 delivery resulted in actual delivery. Cancel cutthrough before DATA
901 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
902 rejects a recipient: pass the reject to the originator.
904 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
905 Many were false-positives and ignorable, but it's worth fixing the
908 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
909 for the new environment-manipulation done at startup. Move the routines
910 from being local to tls.c to being global via the os.c file.
912 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
913 an extract embedded as result-arg for a map, the first arg for extract
914 is unavailable so we cannot tell if this is a numbered or keyed
915 extraction. Accept either.
921 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
924 JH/02 The smtp transport option "multi_domain" is now expanded.
926 JH/03 The smtp transport now requests PRDR by default, if the server offers
929 JH/04 Certificate name checking on server certificates, when exim is a client,
930 is now done by default. The transport option tls_verify_cert_hostnames
931 can be used to disable this per-host. The build option
932 EXPERIMENTAL_CERTNAMES is withdrawn.
934 JH/05 The value of the tls_verify_certificates smtp transport and main options
935 default to the word "system" to access the system default CA bundle.
936 For GnuTLS, only version 3.0.20 or later.
938 JH/06 Verification of the server certificate for a TLS connection is now tried
939 (but not required) by default. The verification status is now logged by
940 default, for both outbound TLS and client-certificate supplying inbound
943 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
946 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
947 Status Notification (bounce) messages are now MIME format per RFC 3464.
948 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
949 under the control of the dsn_advertise_hosts option, and routers may
950 have a dsn_lasthop option.
952 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
953 default, modifiable by a malware= option. The list separator for
954 the options can now be changed in the usual way. Bug 68.
956 JH/10 The smtp_receive_timeout main option is now expanded before use.
958 JH/11 The incoming_interface log option now also enables logging of the
959 local interface on delivery outgoing connections.
961 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
962 if the interface and destination host and port all match.
964 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
967 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
968 Patch from Andrew Lewis.
970 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
971 now supports optional time-restrictions, weighting, and priority
972 modifiers per server. Patch originally by <rommer@active.by>.
974 JH/16 The spamd_address main option now supports a mixed list of local
975 and remote servers. Remote servers can be IPv6 addresses, and
976 specify a port-range.
978 JH/17 Bug 68: The spamd_address main option now supports an optional
979 timeout value per server.
981 JH/18 Bug 1581: Router and transport options headers_add/remove can
982 now have the list separator specified.
984 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
987 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
990 JH/21 Support for the A6 type of dns record is withdrawn.
992 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
993 rather than the verbs used.
995 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
996 from 255 to 1024 chars.
998 JH/24 Verification callouts now attempt to use TLS by default.
1000 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1001 are generic router options now. The defaults didn't change.
1003 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1004 Original patch from Alexander Shikoff, worked over by JH.
1006 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1007 files if ZCAT_COMMAND is not executable.
1009 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1011 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1013 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1014 Normally benign, it bites when the pair was led to by a CNAME;
1015 modern usage is to not canonicalize the domain to a CNAME target
1016 (and we were inconsistent anyway for A-only vs AAAA+A).
1018 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1020 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1021 when evaluating $sender_host_dnssec.
1023 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1024 $sender_helo_dnssec variable.
1026 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1028 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1030 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1032 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1033 documented as working, but never had. Support all but $spam_report.
1035 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1036 added for tls authenticator.
1038 HS/03 Add perl_taintmode main config option
1044 TL/01 When running the test suite, the README says that variables such as
1045 no_msglog_check are global and can be placed anywhere in a specific
1046 test's script, however it was observed that placement needed to be near
1047 the beginning for it to behave that way. Changed the runtest perl
1048 script to read through the entire script once to detect and set these
1049 variables, reset to the beginning of the script, and then run through
1050 the script parsing/test process like normal.
1052 TL/02 The BSD's have an arc4random API. One of the functions to induce
1053 adding randomness was arc4random_stir(), but it has been removed in
1054 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1055 function when detected.
1057 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1058 cause callback expansion.
1060 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1061 syntax errors in an expansion can be treated as a string instead of
1062 logging or causing an error, due to the internal use of bool_lax
1063 instead of bool when processing it.
1065 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1066 server certificates when making smtp deliveries.
1068 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1070 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1072 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1074 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1075 Merged patch from Sebastian Wiedenroth.
1077 JH/05 Fix results-pipe from transport process. Several recipients, combined
1078 with certificate use, exposed issues where response data items split
1079 over buffer boundaries were not parsed properly. This eventually
1080 resulted in duplicates being sent. This issue only became common enough
1081 to notice due to the introduction of connection certificate information,
1082 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1084 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1085 size buffer was used, resulting in syntax errors when an expansion
1088 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1089 version 3.3.6 or later.
1091 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1092 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1093 both become "event_action", the variables become $event_name, $event_data
1094 and $event_defer_errno. There is a new variable $verify_mode, usable in
1095 routers, transports and related events. The tls:cert event is now also
1096 raised for inbound connections, if the main configuration event_action
1099 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1100 early OCSP support, but no stapling (appears to be less than 1.0.0).
1102 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1103 server certificate names available under the smtp transport option
1104 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1107 JH/10 Time-related extraction expansions from certificates now use the main
1108 option "timezone" setting for output formatting, and are consistent
1109 between OpenSSL and GnuTLS compilations. Bug 1541.
1111 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1112 encoded parameter in the incoming message. Bug 1558.
1114 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1115 include certificate info, eximon was claiming there were spoolfile
1118 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1120 JH/14 Log delivery-related information more consistently, using the sequence
1121 "H=<name> [<ip>]" wherever possible.
1123 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1124 are problematic for Debian distribution, omit them from the release
1127 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1129 JH/16 Fix string representation of time values on 64bit time_t architectures.
1132 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1133 output list separator was used.
1138 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1139 checkers that were complaining about end of non-void function with no
1142 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1143 This was a regression introduced in 4.83 by another bugfix.
1145 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1147 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1148 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1154 TF/01 Correctly close the server side of TLS when forking for delivery.
1156 When a message was received over SMTP with TLS, Exim failed to clear up
1157 the incoming connection properly after forking off the child process to
1158 deliver the message. In some situations the subsequent outgoing
1159 delivery connection happened to have the same fd number as the incoming
1160 connection previously had. Exim would try to use TLS and fail, logging
1161 a "Bad file descriptor" error.
1163 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1164 utilities have not been installed.
1166 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1167 temporary space as the ACL may create new global variables.
1169 TL/01 LDAP support uses per connection or global context settings, depending
1170 upon the detected version of the libraries at build time.
1172 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1173 to extract and use the src ip:port in logging and expansions as if it
1174 were a direct connection from the outside internet. PPv2 support was
1175 updated based on HAProxy spec change in May 2014.
1177 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1179 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1180 Properly escape header and check for NULL return.
1182 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1185 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1187 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1188 characters in header names, implemented as a verify condition.
1189 Contributed by Michael Fischer v. Mollard.
1191 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1192 results permerror and temperror. Previous values are deprecated but
1193 still accepted. In a future release, err_perm and err_temp will be
1194 completely removed, which will be a backward incompatibility if the
1195 ACL tests for either of these two old results. Patch contributed by
1196 user bes-internal on the mailing list.
1198 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1200 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1201 selectors, in both main and reject logs.
1203 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1206 JH/07 Add malware type "sock" for talking to simple daemon.
1208 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1210 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1211 routers/transports under cutthrough routing.
1213 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1214 numbers. Touch up "bool" conditional to keep the same definition.
1216 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1218 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1220 JH/12 Expand items in router/transport headers_add or headers_remove lists
1221 individually rather than the list as a whole. Bug 1452.
1223 Required for reasonable handling of multiple headers_ options when
1224 they may be empty; requires that headers_remove items with embedded
1225 colons must have them doubled (or the list-separator changed).
1227 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1228 view the policy declared in the DMARC record. Currently, $dmarc_status
1229 is a combined value of both the record presence and the result of the
1232 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1234 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1235 dnslookup router and the smtp transport (applying to the forward
1238 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1239 of ldap servers used for a specific lookup. Patch provided by Heiko
1242 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1243 New variable $lookup_dnssec_authenticated for observability.
1245 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1246 Patch submitted by Lars Timman.
1248 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1250 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1251 Requires trusted mode and valid format message id, aborts otherwise.
1252 Patch contributed by Heiko Schlichting.
1254 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1255 certextract with support for various fields. Bug 1358.
1257 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1258 is requested by default, modifiable by smtp transport option
1261 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1262 operate on certificate variables to give certificate fingerprints
1263 Also new ${sha256:cert_variable}.
1265 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1267 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1268 Christian Aistleitner.
1270 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1272 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1273 file. Patch from Wolfgang Breyha.
1275 JH/25 Expand the coverage of the delivery $host and $host_address to
1276 client authenticators run in verify callout. Bug 1476.
1278 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1279 align with daemon_smtp_ports. Bug 72.
1281 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1282 support and error reporting did not work properly.
1284 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1285 and is readable. Patch from Andrew Colin Kissa.
1287 TL/14 Enhance documentation of ${run expansion and how it parses the
1288 commandline after expansion, particularly in the case when an
1289 unquoted variable expansion results in an empty value.
1291 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1293 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1294 by renaming away from it. Addresses GH issue 15, reported by
1297 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1299 TL/15 SECURITY: prevent double expansion in math comparison functions
1300 (can expand unsanitized data). Not remotely exploitable.
1307 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1309 PP/02 Make -n do something, by making it not do something.
1310 When combined with -bP, the name of an option is not output.
1312 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1315 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1316 $sender_host_name and config options to manage this, and basic check
1319 PP/05 DSCP support for outbound connections and control modifier for inbound.
1321 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1322 (Only plugin which currently uses this is kerberos4, which nobody should
1323 be using, but we should make it available and other future plugins might
1324 conceivably use it, even though it would break NAT; stuff *should* be
1325 using channel bindings instead).
1327 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1328 name; added for Sendmail compatibility; requires admin caller.
1329 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1330 just ignore it); requires trusted caller.
1331 Also parse but ignore: -Ac -Am -X<logfile>
1334 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1336 TL/02 Add +smtp_confirmation as a default logging option.
1338 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1339 Patch by Magnus Holmgren from 2007-02-20.
1341 TL/04 Bugzilla 1281 - Spec typo.
1342 Bugzilla 1283 - Spec typo.
1343 Bugzilla 1290 - Spec grammar fixes.
1345 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1347 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1349 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1350 mailing list by by Dmitry Isaikin.
1352 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1354 JH/02 Support "G" suffix to numbers in ${if comparisons.
1356 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1358 NM/01 Bugzilla 1197 - Spec typo
1359 Bugzilla 1196 - Spec examples corrections
1361 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1363 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1364 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1367 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1368 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1370 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1371 "acl {{name}{arg}...}", and optional args on acl condition
1374 JH/05 Permit multiple router/transport headers_add/remove lines.
1376 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1378 JH/07 Avoid using a waiting database for a single-message-only transport.
1379 Performance patch from Paul Fisher. Bugzilla 1262.
1381 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1384 JH/09 Add $headers_added variable, with content from use of ACL modifier
1385 add_header (but not yet added to the message). Bugzilla 199.
1387 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1388 Pulled from Bugzilla 817 by Wolfgang Breyha.
1390 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1392 (nb: this is the same fix as in Exim 4.80.1)
1394 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1397 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1400 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1402 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1403 advertise SMTP AUTH mechanism to us, instead of a generic
1404 protocol violation error. Also, make Exim more robust to bad
1405 data from the Dovecot auth socket.
1407 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1409 When a queue runner is handling a message, Exim first routes the
1410 recipient addresses, during which it prunes them based on the retry
1411 hints database. After that it attempts to deliver the message to
1412 any remaining recipients. It then updates the hints database using
1415 So if a recipient address works intermittently, it can get repeatedly
1416 deferred at routing time. The retry hints record remains fresh so the
1417 address never reaches the final cutoff time.
1419 This is a fairly common occurrence when a user is bumping up against
1420 their storage quota. Exim had some logic in its local delivery code
1421 to deal with this. However it did not apply to per-recipient defers
1422 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1424 This change adds a proper retry rule check during routing so that the
1425 final cutoff time is checked against the message's age. We only do
1426 this check if there is an address retry record and there is not a
1427 domain retry record; this implies that previous attempts to handle
1428 the address had the retry_use_local_parts option turned on. We use
1429 this as an approximation for the destination being like a local
1430 delivery, as in LMTP.
1432 I suspect this new check makes the old local delivery cutoff check
1433 redundant, but I have not verified this so I left the code in place.
1435 TF/02 Correct gecos expansion when From: is a prefix of the username.
1437 Test 0254 submits a message to Exim with the header
1441 When I ran the test suite under the user fanf2, Exim expanded
1442 the header to contain my full name, whereas it should have added
1443 a Resent-Sender: header. It erroneously treats any prefix of the
1444 username as equal to the username.
1446 This change corrects that bug.
1448 GF/01 DCC debug and logging tidyup
1449 Error conditions log to paniclog rather than rejectlog.
1450 Debug lines prefixed by "DCC: " to remove any ambiguity.
1452 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1454 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1455 Bug spotted by Jeremy Harris; was flawed since initial commit.
1456 Would have resulted in OCSP responses post-SNI triggering an Exim
1457 NULL dereference and crash.
1459 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1461 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1462 Bug detection, analysis and fix by Samuel Thibault.
1463 Bugzilla 1331, Debian bug #698092.
1465 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1467 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1468 Server implementation by Todd Lyons, client by JH.
1469 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1470 config variable "prdr_enable" controls whether the server
1471 advertises the facility. If the client requests PRDR a new
1472 acl_data_smtp_prdr ACL is called once for each recipient, after
1473 the body content is received and before the acl_smtp_data ACL.
1474 The client is controlled by both of: a hosts_try_prdr option
1475 on the smtp transport, and the server advertisement.
1476 Default client logging of deliveries and rejections involving
1477 PRDR are flagged with the string "PRDR".
1479 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1480 fclose(). Diagnosis by Todd Lyons.
1482 PP/17 Update configure.default to handle IPv6 localhost better.
1483 Patch by Alain Williams (plus minor tweaks).
1486 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1487 This is now consistent with GnuTLS, and is now documented: the
1488 previous undocumented portable approach to treating the option as
1489 unset was to force an expansion failure. That still works, and
1490 an empty string is now equivalent.
1492 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1493 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1494 not performing validation itself.
1496 PP/20 Added force_command boolean option to pipe transport.
1497 Patch from Nick Koston, of cPanel Inc.
1499 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1502 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1504 PP/21 Fix eximon continuous updating with timestamped log-files.
1505 Broken in a format-string cleanup in 4.80, missed when I repaired the
1506 other false fix of the same issue.
1507 Report and fix from Heiko Schlichting.
1510 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1511 Report from Prashanth Katuri.
1513 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1514 It's SecureTransport, so affects any MacOS clients which use the
1515 system-integrated TLS libraries, including email clients.
1517 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1518 using a MIME ACL for non-SMTP local injection.
1519 Report and assistance in diagnosis by Warren Baker.
1521 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1523 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1525 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1526 last id that failed so it may be referenced in subsequent ACL's.
1528 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1531 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1532 ldap library initialization, allowing self-signed CA's to be
1533 used. Also properly sets require_cert option later in code by
1534 using NULL (global ldap config) instead of ldap handle (per
1535 session). Bug diagnosis and testing by alxgomz.
1537 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1538 the src/util/ subdirectory.
1540 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1541 renamed to Transport Post Delivery Action by Jeremy Harris, as
1544 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1545 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1546 redis_servers = needs to be configured which will be used by the redis
1547 lookup. Patch from Warren Baker, of The Packet Hub.
1549 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1551 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1552 hostname or reverse DNS when processing a host list. Used suggestions
1553 from multiple comments on this bug.
1555 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1557 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1558 Missed a few lines, added it to make the runtest require no keyboard
1561 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1562 contains upper case chars. Make router use caseful_local_part.
1564 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1565 support when GnuTLS has been built with p11-kit.
1571 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1573 This, or similar/improved, will also be change PP/11 of 4.82.
1579 PP/01 Handle short writes when writing local log-files.
1580 In practice, only affects FreeBSD (8 onwards).
1581 Bugzilla 1053, with thanks to Dmitry Isaikin.
1583 NM/01 Bugzilla 949 - Documentation tweak
1585 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1588 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1590 PP/02 Implemented gsasl authenticator.
1592 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1594 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1595 `pkg-config foo` for cflags/libs.
1597 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1598 with rest of GSASL and with heimdal_gssapi.
1600 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1601 `pkg-config foo` for cflags/libs for the TLS implementation.
1603 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1604 properties get this fed in as external SSF. A number of robustness
1605 and debugging improvements to the cyrus_sasl authenticator.
1607 PP/08 cyrus_sasl server now expands the server_realm option.
1609 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1610 Patch by Jeremy Harris.
1612 PP/10 Added dbmjz lookup type.
1614 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1616 PP/12 MAIL args handles TAB as well as SP, for better interop with
1617 non-compliant senders.
1618 Analysis and variant patch by Todd Lyons.
1620 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1621 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1622 Patch from Dirk Mueller <dmueller@suse.com>
1624 PP/13 tls_peerdn now print-escaped for spool files.
1625 Observed some $tls_peerdn in wild which contained \n, which resulted
1626 in spool file corruption.
1628 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1629 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1630 or write after TLS renegotiation, which otherwise led to messages
1633 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1634 as a tracking header (ie: a signed header comes before the signature).
1635 Patch from Wolfgang Breyha.
1637 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1638 comma-sep list; embedded commas doubled.
1640 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1642 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1644 Report and patch from Dmitry Banschikov.
1646 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1647 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1648 failures. We appear to now support TLS1.1+ with Exim.
1650 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1651 lets Exim select keys and certificates based upon TLS SNI from client.
1652 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1653 before an outbound SMTP session. New log_selector, +tls_sni.
1655 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1656 NULL dereference. Report and patch from Alun Jones.
1658 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1659 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1660 Not seeing resolver debug output on NetBSD, but suspect this is a
1661 resolver implementation change.
1663 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1664 Left warnings. Added "eximon gdb" invocation mode.
1666 PP/21 Defaulting "accept_8bitmime" to true, not false.
1668 PP/22 Added -bw for inetd wait mode support.
1670 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1671 locate the relevant includes and libraries. Made this the default.
1673 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1674 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1676 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1677 This may cause build issues on older platforms.
1679 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1680 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1681 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1682 Added SNI support via GnuTLS too.
1683 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1685 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1687 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1688 Applied second patch from Janne, implementing suggestion to default
1689 multiple-strings-in-record handling to match SPF spec.
1691 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1693 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1694 read-only, out of scope).
1695 Patch from Wolfgang Breyha, report from Stuart Northfield.
1697 PP/29 Fix three issues highlighted by clang analyser static analysis.
1698 Only crash-plausible issue would require the Cambridge-specific
1699 iplookup router and a misconfiguration.
1700 Report from Marcin Mirosław.
1702 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1704 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1705 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1706 As part of this, removing so much warning spew let me fix some minor
1707 real issues in debug logging.
1709 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1710 assignment on my part. Fixed.
1712 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1713 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1714 Janne Snabb (who went above and beyond: thank you).
1716 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1717 string otherwise requires a connection and a bunch more work and it's
1718 relatively easy to get wrong. Should also expose TLS library linkage
1721 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1722 64-bit ${eval} (JH/03).
1724 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1725 GNU libc to support some of the 64-bit stuff, should not lead to
1726 conflicts. Defined before os.h is pulled in, so if a given platform
1727 needs to override this, it can.
1729 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1730 protection layer was required, which is not implemented.
1731 Bugzilla 1254, patch from Wolfgang Breyha.
1733 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1734 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1735 tls_dhparam take prime identifiers. Also unbreak combination of
1736 OpenSSL+DH_params+TLSSNI.
1738 PP/39 Disable SSLv2 by default in OpenSSL support.
1744 PP/01 Solaris build fix for Oracle's LDAP libraries.
1745 Bugzilla 1109, patch from Stephen Usher.
1747 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1749 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1752 TF/02 Fix a couple more cases where we did not log the error message
1753 when unlink() failed. See also change 4.74-TF/03.
1755 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1756 lock up or crash if it happened to be inside a call to libc when it
1757 got a SIGUSR1 from exiwhat.
1759 The SIGUSR1 handler appends the current process status to the process
1760 log which is later printed by exiwhat. It used to use the general
1761 purpose logging code to do this, but several functions it calls are
1762 not safe for signals.
1764 The new output code in the SIGUSR1 handler is specific to the process
1765 log, and simple enough that it's easy to inspect for signal safety.
1766 Removing some special cases also simplifies the general logging code.
1767 Removing the spurious timestamps from the process log simplifies
1770 TF/04 Improved ratelimit ACL condition.
1772 The /noupdate option has been deprecated in favour of /readonly which
1773 has clearer semantics. The /leaky, /strict, and /readonly update modes
1774 are mutually exclusive. The update mode is no longer included in the
1775 database key; it just determines when the database is updated. (This
1776 means that when you upgrade Exim will forget old rate measurements.)
1778 Exim now checks that the per_* options are used with an update mode that
1779 makes sense for the current ACL. For example, when Exim is processing a
1780 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1781 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1782 must specify per_mail/readonly. If you omit the update mode it defaults to
1783 /leaky where that makes sense (as before) or /readonly where required.
1785 The /noupdate option is now undocumented but still supported for
1786 backwards compatibility. It is equivalent to /readonly except that in
1787 ACLs where /readonly is required you may specify /leaky/noupdate or
1788 /strict/noupdate which are treated the same as /readonly.
1790 A useful new feature is the /count= option. This is a generalization
1791 of the per_byte option, so that you can measure the throughput of other
1792 aggregate values. For example, the per_byte option is now equivalent
1793 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1795 The per_rcpt option has been generalized using the /count= mechanism
1796 (though it's more complicated than the per_byte equivalence). When it is
1797 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1798 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1799 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1800 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1801 Note that using per_rcpt with a non-readonly update mode in more than
1802 one ACL will cause the recipients to be double-counted. (The per_mail
1803 and per_byte options don't have this problem.)
1805 The handling of very low rates has changed slightly. If the computed rate
1806 is less than the event's count (usually one) then this event is the first
1807 after a long gap. In this case the rate is set to the same as this event's
1808 count, so that the first message of a spam run is counted properly.
1810 The major new feature is a mechanism for counting the rate of unique
1811 events. The new per_addr option counts the number of different
1812 recipients that someone has sent messages to in the last time period. It
1813 behaves like per_rcpt if all the recipient addresses are different, but
1814 duplicate recipient addresses do not increase the measured rate. Like
1815 the /count= option this is a general mechanism, so the per_addr option
1816 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1817 example, measure the rate that a client uses different sender addresses
1818 with the options per_mail/unique=$sender_address. There are further
1819 details in the main documentation.
1821 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1823 TF/06 Removed a few PCRE remnants.
1825 TF/07 Automatically extract Exim's version number from tags in the git
1826 repository when doing development or release builds.
1828 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1829 Bugzilla 879. Patch from Paul Fisher.
1831 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1832 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1835 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1837 PP/05 Variable $av_failed, true if the AV scanner deferred.
1838 Bugzilla 1078. Patch from John Horne.
1840 PP/06 Stop make process more reliably on build failure.
1841 Bugzilla 1087. Patch from Heiko Schlittermann.
1843 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1844 Bugzilla 1089. Patch from Heiko Schlittermann.
1846 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1847 Bugzilla 1131. Patch from Holger Weiß.
1849 PP/09 Handle IPv6 addresses with SPF.
1850 Bugzilla 860. Patch from Wolfgang Breyha.
1852 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1854 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1857 PP/11 match_* no longer expand right-hand-side by default.
1858 New compile-time build option, EXPAND_LISTMATCH_RHS.
1859 New expansion conditions, "inlist", "inlisti".
1861 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1863 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1865 PP/14 fix log_write() format string regression from TF/03.
1866 Bugzilla 1152. Patch from Dmitry Isaikin.
1872 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1874 PP/02 Harmonised TLS library version reporting; only show if debugging.
1875 Layout now matches that introduced for other libraries in 4.74 PP/03.
1877 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1879 PP/04 New "dns_use_edns0" global option.
1881 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1884 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1885 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1887 TK/01 Updated PolarSSL code to 0.14.2.
1888 Bugzilla 1097. Patch from Andreas Metzler.
1890 PP/07 Catch divide-by-zero in ${eval:...}.
1891 Fixes bugzilla 1102.
1893 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1896 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1897 format-string attack -- SECURITY: remote arbitrary code execution.
1899 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1900 time unintentionally subject to list matching rules, letting the header
1901 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1902 arbitrary string expansion). This allowed for information disclosure.
1904 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1905 INT_MIN/-1 -- value coerced to INT_MAX.
1911 NM/01 Workaround for PCRE version dependency in version reporting
1914 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1915 This fixes portability to compilers other than gcc, notably
1916 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1918 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1919 makefiles for portability to HP-UX and POSIX correctness.
1921 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1922 Also via indented variable definition in the Makefile.
1923 (Debugging by Oliver Heesakkers).
1925 PP/02 Restore caching of spamd results with expanded spamd_address.
1926 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1928 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1929 Improves build reliability. Fix from: Frank Elsner
1931 NM/02 Fix wide character breakage in the rfc2047 coding
1932 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1934 NM/03 Allow underscore in dnslist lookups
1935 Fixes bug 1026. Patch from Graeme Fowler
1937 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1938 Code patches from Adam Ciarcinski of NetBSD.
1940 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1943 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1944 is logged, to avoid truncation. Patch from John Horne.
1946 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1947 Patch from Jakob Hirsch.
1949 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1950 SQL string expansion failure details.
1951 Patch from Andrey Oktyabrski.
1953 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1954 Patch from Simon Arlott.
1956 PP/09 New lookups functionality failed to compile on old gcc which rejects
1957 extern declarations in function scope.
1958 Patch from Oliver Fleischmann
1960 PP/10 Use sig_atomic_t for flags set from signal handlers.
1961 Check getgroups() return and improve debugging.
1962 Fixed developed for diagnosis in bug 927 (which turned out to be
1965 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1966 Patch from Mark Zealey.
1968 PP/12 Bugzilla 1056: Improved spamd server selection.
1969 Patch from Mark Zealey.
1971 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1972 Based on patch from Heiko Schlittermann.
1974 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1975 Patch from Uwe Doering, sign-off by Michael Haardt.
1977 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1978 variable declaration deep within a block. Bug and patch from
1981 PP/15 lookups-Makefile IRIX compatibility coercion.
1983 PP/16 Make DISABLE_DKIM build knob functional.
1985 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1986 Patch by Simon Arlott
1988 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1989 variable argument macros. Our copy now differs from upstream.
1995 TF/01 Failure to get a lock on a hints database can have serious
1996 consequences so log it to the panic log.
1998 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1999 controlled using the smtp_confirmation log selector.
2001 TF/03 Include the error message when we fail to unlink a spool file.
2003 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2004 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2005 for maintaining out-of-tree patches for some time.
2007 PP/01 Bugzilla 139: Documentation and portability issues.
2008 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2009 Handle per-OS dynamic-module compilation flags.
2011 PP/02 Let /dev/null have normal permissions.
2012 The 4.73 fixes were a little too stringent and complained about the
2013 permissions on /dev/null. Exempt it from some checks.
2014 Reported by Andreas M. Kirchwitz.
2016 PP/03 Report version information for many libraries, including
2017 Exim version information for dynamically loaded libraries. Created
2018 version.h, now support a version extension string for distributors
2019 who patch heavily. Dynamic module ABI change.
2021 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2022 privilege escalation vulnerability whereby the Exim run-time user
2023 can cause root to append content of the attacker's choosing to
2026 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2029 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2030 If dropping privileges for untrusted macros, we disabled normal logging
2031 on the basis that it would fail; for the Exim run-time user, this is not
2032 the case, and it resulted in successful deliveries going unlogged.
2033 Fixed. Reported by Andreas Metzler.
2039 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2040 only prepend for the Resent-* case. Fixes regression introduced in
2041 Exim 4.70 by NM/22 for Bugzilla 607.
2043 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2044 increasing numbers of administrators be bitten by this.
2046 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2048 PP/03 Bugzilla 994: added openssl_options main configuration option.
2050 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2052 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2054 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2056 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2057 without a peer certificate, leading to a segfault because of an
2058 assumption that peers always have certificates. Be a little more
2059 paranoid. Problem reported by Martin Tscholak.
2061 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2062 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2063 NB: ClamAV planning to remove STREAM in "middle of 2010".
2064 CL also introduces -bmalware, various -d+acl logging additions and
2065 more caution in buffer sizes.
2067 PP/09 Implemented reverse_ip expansion operator.
2069 PP/10 Bugzilla 937: provide a "debug" ACL control.
2071 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2073 PP/12 Bugzilla 973: Implement --version.
2075 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2077 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2079 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2081 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2082 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2083 ignore trailing whitespace.
2085 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2087 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2090 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2091 Notification from Dr Andrew Aitchison.
2093 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2094 ExtendedDetectionInfo response format.
2095 Notification from John Horne.
2097 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2100 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2101 XSL and documented dependency on system catalogs, with examples of how
2104 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2107 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2108 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2109 configuration file which is writeable by the Exim user or group.
2111 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2112 of configuration files to cover files specified with the -C option if
2113 they are going to be used with root privileges, not just the default
2116 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2117 option (effectively making it always true).
2119 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2120 files to be used while preserving root privileges.
2122 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2123 that rogue child processes cannot use them.
2125 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2126 run-time user, instead of root.
2128 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2129 Exim run-time user without dropping privileges.
2131 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2132 result string, instead of calling string_vformat() twice with the same
2135 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2136 for other users. Others should always drop root privileges if they use
2137 -C on the command line, even for a whitelisted configure file.
2139 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2141 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2147 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2148 $data_path, and $header_path variables; fixed documentation bugs and
2151 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2152 exipick to access non-standard spools, including the "frozen" queue
2155 NM/01 Bugzilla 965: Support mysql stored procedures.
2156 Patch from Alain Williams
2158 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2160 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2161 Patch from Andreas Metzler
2163 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2164 Patch from Kirill Miazine
2166 NM/05 Bugzilla 671: Added umask to procmail example.
2168 JJ/03 installed exipick 20100323.0, fixing doc bug
2170 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2171 directory. Notification and patch from Dan Rosenberg.
2173 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2175 TK/02 Improve log output when DKIM signing operation fails.
2177 MH/01 Treat the transport option dkim_domain as a colon separated
2178 list, not as a single string, and sign the message with each element,
2179 omitting multiple occurences of the same signer.
2181 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2182 Bugzilla 985, 986. Patch by Simon Arlott
2184 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2185 Patch by Simon Arlott
2187 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2188 MBX locking. Notification from Dan Rosenberg.
2194 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2196 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2198 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2200 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2202 NM/04 Bugzilla 915: Flag broken perl installation during build.
2208 TK/01 Added patch by Johannes Berg that expands the main option
2209 "spamd_address" if it starts with a dollar sign.
2211 TK/02 Write list of recipients to X-Envelope-Sender header when building
2212 the mbox-format spool file for content scanning (suggested by Jakob
2215 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2216 (http://www.dcc-servers.net/) support via dccifd. Activated by
2217 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2219 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2220 by Mark Daniel Reidel <mr@df.eu>.
2222 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2223 When building exim an external PCRE library is now needed -
2224 PCRE is a system library on the majority of modern systems.
2225 See entry on PCRE_LIBS in EDITME file.
2227 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2228 conversation. Added nologin parameter to request.
2229 Patch contributed by Kirill Miazine.
2231 TF/01 Do not log submission mode rewrites if they do not change the address.
2233 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2235 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2236 log files in place. Contributed by Roberto Lima.
2238 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2240 TF/03 Bugzilla 615: When checking the local_parts router precondition
2241 after a local_part_suffix or local_part_prefix option, Exim now
2242 does not use the address's named list lookup cache, since this
2243 contains cached lookups for the whole local part.
2245 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2246 Robert Millan. Documentation is in experimental-spec.txt.
2248 TF/04 Bugzilla 668: Fix parallel build (make -j).
2250 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2252 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2253 Patch provided by Jan Srzednicki.
2255 TF/05 Leading white space used to be stripped from $spam_report which
2256 wrecked the formatting. Now it is preserved.
2258 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2259 that they are available at delivery time.
2261 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2263 TF/08 TLS error reporting now respects the incoming_interface and
2264 incoming_port log selectors.
2266 TF/09 Produce a more useful error message if an SMTP transport's hosts
2267 setting expands to an empty string.
2269 NM/06 Bugzilla 744: EXPN did not work under TLS.
2270 Patch provided by Phil Pennock.
2272 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2273 Patch provided by Richard Godbee.
2275 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2276 acl_smtp_notquit, added index entry.
2278 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2279 Patch provided by Eugene Bujak.
2281 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2282 accept(). Patch provided by Maxim Dounin.
2284 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2285 Patch provided by Phil Pennock.
2287 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2289 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2290 Patch provided by Brad "anomie" Jorsch.
2292 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2293 Patch provided by Dean Brooks.
2295 TK/05 Add native DKIM support (does not depend on external libraries).
2297 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2298 Patch provided by Graeme Fowler.
2300 NM/16 Bugzilla 851: Documentation example syntax fix.
2302 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2304 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2307 NM/19 Bugzilla 745: TLS version reporting.
2308 Patch provided by Phil Pennock.
2310 NM/20 Bugzilla 167: bool: condition support.
2311 Patch provided by Phil Pennock.
2313 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2314 clients. Patch provided by Phil Pennock.
2316 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2317 Patch provided by Brad "anomie" Jorsch.
2319 NM/23 Bugzilla 687: Fix misparses in eximstats.
2320 Patch provided by Heiko Schlittermann.
2322 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2323 Patch provided by Heiko Schlittermann.
2325 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2326 plus update to original patch.
2328 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2330 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2331 Patch provided by David Brownlee.
2333 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2335 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2337 NM/30 Bugzilla 888: TLS documentation bugfixes.
2339 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2341 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2342 Unlike the original bugzilla I have changed all shell scripts in src tree.
2344 NM/33 Bugzilla 898: Transport filter timeout fix.
2345 Patch by Todd Rinaldo.
2347 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2348 Patch by Serge Demonchaux.
2350 NM/35 Bugzilla 39: Base64 decode bug fixes.
2351 Patch by Jakob Hirsch.
2353 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2355 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2357 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2359 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2365 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2366 ALT-N's libdkim that I have put here:
2367 http://duncanthrax.net/exim-experimental/
2369 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2370 were called 'true' and it seems that C99 defines that as a reserved
2371 keyword to be used with 'bool' variable types. That means you could
2372 not include C99-style headers which use bools without triggering
2373 build errors in sieve.c.
2375 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2376 as mailq or other aliases. Changed the --help handling significantly
2377 to do whats expected. exim_usage() emits usage/help information.
2379 SC/01 Added the -bylocaldomain option to eximstats.
2381 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2383 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2385 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2391 PH/01 Another patch from the Sieve maintainer.
2393 PH/02 When an IPv6 address is converted to a string for single-key lookup
2394 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2395 dots are used instead of colons so that keys in lsearch files need not
2396 contain colons. This was done some time before quoting was made available
2397 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2398 (notated using the quote facility) so as to distinguish them from IPv4
2399 keys. This meant that lookups for IP addresses in host lists did not work
2400 for iplsearch lookups.
2402 This has been fixed by arranging for IPv6 addresses to be expressed with
2403 colons if the lookup type is iplsearch. This is not incompatible, because
2404 previously such lookups could never work.
2406 The situation is now rather anomalous, since one *can* have colons in
2407 ordinary lsearch keys. However, making the change in all cases is
2408 incompatible and would probably break a number of configurations.
2410 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2413 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2414 conversion specification without a maximum field width, thereby enabling
2415 a rogue spamd server to cause a buffer overflow. While nobody in their
2416 right mind would setup Exim to query an untrusted spamd server, an
2417 attacker that gains access to a server running spamd could potentially
2418 exploit this vulnerability to run arbitrary code as the Exim user.
2420 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2421 $primary_hostname instead of what libspf2 thinks the hosts name is.
2423 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2424 a directory entry by the name of the lookup key. Previously, if a
2425 symlink pointed to a non-existing file or a file in a directory that
2426 Exim lacked permissions to read, a lookup for a key matching that
2427 symlink would fail. Now it is enough that a matching directory entry
2428 exists, symlink or not. (Bugzilla 503.)
2430 PH/03 The body_linecount and body_zerocount variables are now exported in the
2433 PH/04 Added the $dnslist_matched variable.
2435 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2436 This means they are set thereafter only if the connection becomes
2439 PH/06 Added the client_condition to authenticators so that some can be skipped
2440 by clients under certain conditions.
2442 PH/07 The error message for a badly-placed control=no_multiline_responses left
2443 "_responses" off the end of the name.
2445 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2447 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2448 (without spaces) instead of just copying the configuration text.
2450 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2452 PH/11 Added $max_received_linelength.
2454 PH/12 Added +ignore_defer and +include_defer to host lists.
2456 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2457 way in which PCRE > 7.0 is built.
2459 PH/14 Implemented queue_only_load_latch.
2461 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2462 MAIL command. The effect was to mangle the value on 64-bit systems.
2464 PH/16 Another patch from the Sieve maintainer.
2466 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2468 PH/18 If a system quota error occurred while trying to create the file for
2469 a maildir delivery, the message "Mailbox is full" was not appended to the
2470 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2471 only to a quota excession during the actual writing of the file.
2473 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2474 characters?) which causes problems in log lines. The DN values are now
2475 passed through string_printing() before being added to log lines.
2477 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2478 and InterBase are left for another time.)
2480 PH/21 Added message_body_newlines option.
2482 PH/22 Guard against possible overflow in moan_check_errorcopy().
2484 PH/23 POSIX allows open() to be a macro; guard against that.
2486 PH/24 If the recipient of an error message contained an @ in the local part
2487 (suitably quoted, of course), incorrect values were put in $domain and
2488 $local_part during the evaluation of errors_copy.
2494 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2495 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2498 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2499 issue a MAIL command.
2501 PH/02 In an ACL statement such as
2503 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2505 if a client was not listed at all, or was listed with a value other than
2506 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2507 the condition was not true (as it should be), so access was not denied.
2508 The bug was that the ! inversion was incorrectly passed on to the second
2509 item. This has been fixed.
2511 PH/03 Added additional dnslists conditions == and =& which are different from
2512 = and & when the dns lookup returns more than one IP address.
2514 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2515 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2517 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2518 FSYNC, which compiles an option called disable_fsync that allows for
2519 bypassing fsync(). The documentation is heavily laced with warnings.
2521 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2523 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2524 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2525 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2526 including adding "make clean"; (3) Added -fPIC when compiling the test
2527 dynamically loaded module, to get rid of a warning.
2529 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2530 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2531 = 0s. The bug is otherwise harmless.
2533 PH/07 There was a bug in the dovecot authenticator such that the value of
2534 $auth1 could be overwritten, and so not correctly preserved, after a
2535 successful authentication. This usually meant that the value preserved by
2536 the server_setid option was incorrect.
2538 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2540 PH/09 Installed PCRE release 7.0.
2542 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2543 run for batched SMTP input. It is now run at the start of every message
2544 in the batch. While fixing this I discovered that the process information
2545 (output by running exiwhat) was not always getting set for -bs and -bS
2546 input. This is fixed, and it now also says "batched" for BSMTP.
2548 PH/11 Added control=no_pipelining.
2550 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2551 patch, slightly modified), and move the expansion of helo_data till after
2552 the connection is made in the smtp transport (so it can use these
2555 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2557 PH/14 Added log_selector = +pid.
2559 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2561 PH/16 Add ${if forany and ${if forall.
2563 PH/17 Added dsn_from option to vary the From: line in DSNs.
2565 PH/18 Flush SMTP output before performing a callout, unless control =
2566 no_callout_flush is set.
2568 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2569 was true (the default) a successful delivery failed to delete the retry
2570 item, thus causing premature timeout of the address. The bug is now
2573 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2575 PH/21 Long custom messages for fakedefer and fakereject are now split up
2576 into multiline responses in the same way that messages for "deny" and
2577 other ACL rejections are.
2579 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2580 with slight modification.
2582 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2583 draft, changing the syntax and factoring some duplicate code".
2585 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2586 for deliveries of the second and subsequent messages over the same SMTP
2589 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2590 ${reduce, with only minor "tidies".
2592 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2594 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2595 expansion side effects.
2597 PH/27 When a message times out after an over-quota error from an Exim-imposed
2598 quota, the bounce message says "mailbox is full". This message was not
2599 being given when it was a system quota that was exceeded. It now should
2602 MH/03 Made $recipients available in local_scan(). local_scan() already has
2603 better access to the recipient list through recipients_list[], but
2604 $recipients can be useful in postmaster-provided expansion strings.
2606 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2607 in the case of a MAIL command with additional options following the
2608 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2609 were accidentally chopped off.
2611 PH/29 SMTP synchronization checks are implemented when a command is read -
2612 there is a check that no more input is waiting when there shouldn't be
2613 any. However, for some commands, a delay in an ACL can mean that it is
2614 some time before the response is written. In this time, more input might
2615 arrive, invalidly. So now there are extra checks after an ACL has run for
2616 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2617 pipelining has not been advertised.
2619 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2621 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2622 returned in the SMTP response when smtp_return_error_details was false.
2623 This has been fixed.
2625 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2626 instead of the C I/O that was originally supplied, because problems were
2627 reported on Solaris.
2629 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2630 Exim which did not show up earlier: it was assuming that a call to
2631 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2632 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2633 was a macro that became an assignment, so it seemed to work. This has
2634 changed to a proper function call with a void return, hence the compile
2635 error. Exim's code has been fixed.
2637 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2640 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2642 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2644 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2645 because it turns out to be too big or there is a timeout) there is no
2646 "Completed" line in the log. When some messages of this type were
2647 selected by exigrep, they were listed as "not completed". Others were
2648 picked up by some special patterns. I have improved the selection
2649 criteria to be more general.
2651 PH/38 The host_find_failed option in the manualroute router can now be set
2652 to "ignore", to completely ignore a host whose IP address cannot be
2653 found. If all hosts are ignored, the behaviour is controlled by the new
2654 host_all_ignored option.
2656 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2657 homing or because of multiple MX records with /mx) generated more than
2658 one IP address, and the following item turned out to be the local host,
2659 all the secondary addresses of the first item were incorrectly removed
2660 from the list, along with the local host and any following hosts (which
2661 is what is supposed to happen).
2663 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2664 whoever called Exim into the -H file. In the case of the daemon it was
2665 behaving confusingly. When first started, it used values for whoever
2666 started the daemon, but after a SIGHUP it used the Exim user (because it
2667 calls itself on a restart). I have changed the code so that it now always
2670 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2671 message are rejected with the same error (e.g. no authentication or bad
2672 sender address), and a DATA command is nevertheless sent (as can happen
2673 with PIPELINING or a stupid MUA), the error message that was given to the
2674 RCPT commands is included in the rejection of the DATA command. This is
2675 intended to be helpful for MUAs that show only the final error to their
2678 PH/42 Another patch from the Sieve maintainer.
2680 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2681 Eximstats - Fixed some broken HTML links and added missing column headers
2683 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2684 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2686 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2688 PH/43 Yet another patch from the Sieve maintainer.
2690 PH/44 I found a way to check for a TCP/IP connection going away before sending
2691 the response to the final '.' that terminates a message, but only in the
2692 case where the client has not sent further data following the '.'
2693 (unfortunately, this is allowed). However, in many cases there won't be
2694 any further data because there won't be any more messages to send. A call
2695 to select() can be used: if it shows that the input is "ready", there is
2696 either input waiting, or the socket has been closed. An attempt to read
2697 the next input character can distinguish the two cases. Previously, Exim
2698 would have sent an OK response which the client would never have see.
2699 This could lead to message repetition. This fix should cure that, at
2700 least in a lot of common cases.
2702 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2703 advertised in response to EHLO.
2709 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2710 fixed by 4.65/MH/01 (is this a record?) are fixed:
2712 (i) An empty string was always treated as zero by the numeric comparison
2713 operators. This behaviour has been restored.
2715 (ii) It is documented that the numeric comparison operators always treat
2716 their arguments as decimal numbers. This was broken in that numbers
2717 starting with 0 were being interpreted as octal.
2719 While fixing these problems I realized that there was another issue that
2720 hadn't been noticed. Values of message_size_limit (both the global option
2721 and the transport option) were treated as octal if they started with 0.
2722 The documentation was vague. These values are now always treated as
2723 decimal, and I will make that clear in the documentation.
2729 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2730 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2733 MH/01 Don't check that the operands of numeric comparison operators are
2734 integers when their expansion is in "skipping" mode (fixes bug
2735 introduced by 4.64-PH/07).
2737 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2738 child addresses, Exim now panics and dies. Previously, because the count
2739 is held in a short int, deliveries were likely to be lost. As such a
2740 large number of recipients for a single message is ridiculous
2741 (performance will be very, very poor), I have chosen to impose a limit
2742 rather than extend the field.
2748 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2749 leftover -K file (the existence of which was triggered by #402).
2750 While we were at it, introduced process PID as part of the -K
2751 filename. This should rule out race conditions when creating
2754 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2755 processing considerably. Previous code took too long for large mails,
2756 triggering a timeout which in turn triggers #401.
2758 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2759 in the DK code in transports.c. sendfile() is not really portable,
2760 hence the _LINUX specificness.
2762 TF/01 In the add_headers option to the mail command in an Exim filter,
2763 there was a bug that Exim would claim a syntax error in any
2764 header after the first one which had an odd number of characters
2767 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2768 callout verification, Exim cached a "reject" for the entire domain. This
2769 is correct for most verifications, but it is not correct for a recipient
2770 verification with use_sender or use_postmaster set, because in that case
2771 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2772 case of MAIL FROM:<> rejection from other early rejections (e.g.
2773 rejection of HELO). When verifying a recipient using a non-null MAIL
2774 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2775 Whatever the result of the callout, the value of the domain cache is
2776 left unchanged (for any other kind of callout, getting as far as trying
2777 RCPT means that the domain itself is ok).
2779 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2782 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2783 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2784 session, the daemon ignores ECONNECT errors and logs others; it now
2785 ignores EPIPE as well.
2787 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2788 (quoted-printable decoding).
2790 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2791 later the small subsequent patch to fix an introduced bug.
2793 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2795 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2797 PH/08 An error is now given if message_size_limit is specified negative.
2799 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2800 to be given (somewhat) arbitrary names.
2802 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2805 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2806 miscellaneous code fixes
2808 PH/10 Added the log_reject_target ACL modifier to specify where to log
2811 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2812 hostname. This is wrong, because it relates to the incoming message (and
2813 probably the interface on which it is arriving) and not to the outgoing
2814 callout (which could be using a different interface). This has been
2815 changed to use the value of the helo_data option from the smtp transport
2816 instead - this is what is used when a message is actually being sent. If
2817 there is no remote transport (possible with a router that sets up host
2818 addresses), $smtp_active_hostname is used.
2820 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2821 tweaks were necessary in order to get it to work (see also 21 below):
2822 (a) The code assumed that strncpy() returns a negative number on buffer
2823 overflow, which isn't the case. Replaced with Exim's string_format()
2825 (b) There were several signed/unsigned issues. I just did the minimum
2826 hacking in of casts. There is scope for a larger refactoring.
2827 (c) The code used strcasecmp() which is not a standard C function.
2828 Replaced with Exim's strcmpic() function.
2829 (d) The code set only $1; it now sets $auth1 as well.
2830 (e) A simple test gave the error "authentication client didn't specify
2831 service in request". It would seem that Dovecot has changed its
2832 interface. Fortunately there's a specification; I followed it and
2833 changed what the client sends and it appears to be working now.
2835 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2838 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2839 address A is aliased to B and C, where B exists and C does not. Without
2840 -v the output is "A verified" because verification stops after a
2841 successful redirection if more than one address is generated. However,
2842 with -v the child addresses are also verified. Exim was outputting "A
2843 failed to verify" and then showing the successful verification for C,
2844 with its parentage. It now outputs "B failed to verify", showing B's
2845 parentage before showing the successful verification of C.
2847 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2848 look up a TXT record in a specific list after matching in a combined
2851 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2852 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2853 they consult the DNS. I had assumed they would set it the way they
2854 wanted; and indeed my experiments on Linux seem to show that in some
2855 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2856 To be on the safe side, however, I have now made the interface to
2857 host_find_byname() similar to host_find_bydns(), with an argument
2858 containing the DNS resolver options. The host_find_byname() function now
2859 sets these options at its start, just as host_find_bydns() does. The smtp
2860 transport options dns_qualify_single and dns_search_parents are passed to
2861 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2862 of host_find_byname() use the default settings of RES_DEFNAMES
2863 (qualify_single) but not RES_DNSRCH (search_parents).
2865 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2866 spool_read_header() do less string testing, by means of a preliminary
2867 switch on the second character of optional "-foo" lines. (This is
2868 overdue, caused by the large number of possibilities that now exist.
2869 Originally there were few.) While I was there, I also converted the
2870 str(n)cmp tests so they don't re-test the leading "-" and the first
2871 character, in the hope this might squeeze out yet more improvement.
2873 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2874 flag allowing group syntax was set by the header_syntax check but not
2875 turned off, possible causing trouble later; (2) The flag was not being
2876 set at all for the header_verify test, causing "group"-style headers to
2877 be rejected. I have now set it in this case, and also caused header_
2878 verify to ignore an empty address taken from a group. While doing this, I
2879 came across some other cases where the code for allowing group syntax
2880 while scanning a header line wasn't quite right (mostly, not resetting
2881 the flag correctly in the right place). These bugs could have caused
2882 trouble for malformed header lines. I hope it is now all correct.
2884 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2885 with the "reply" argument non-NULL. The code, however (which originally
2886 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2887 but it didn't always do it. This confused somebody who was copying the
2888 code for some other use. I have removed all the tests.
2890 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2891 feature that was used to support insecure browsers during the U.S. crypto
2892 embargo. It requires special client support, and Exim is probably the
2893 only MTA that supported it -- and would never use it because real RSA is
2894 always available. This code has been removed, because it had the bad
2895 effect of slowing Exim down by computing (never used) parameters for the
2896 RSA_EXPORT functionality.
2898 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2899 authenticator to fail if there's a tab character in the incoming data
2900 (there should never be unless someone is messing about, as it's supposed
2901 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2902 if the connection is using TLS or if the remote IP is the same as the
2903 local IP, and the "valid-client-cert option" if a client certificate has
2906 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2907 authenticators. This can be used for authorization after authentication
2908 succeeds. (In the case of plaintext, it servers for both authentication
2911 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2912 if any retry times were supplied.
2914 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2915 connection, where there is no client IP address to check. In this
2916 situation, the verify now always succeeds.
2918 PH/25 Applied John Jetmore's -Mset patch.
2920 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2922 PH/27 In a string expansion for a processed (not raw) header when multiple
2923 headers of the same name were present, leading whitespace was being
2924 removed from all of them, but trailing whitespace was being removed only
2925 from the last one. Now trailing whitespace is removed from each header
2926 before concatenation. Completely empty headers in a concatenation (as
2927 before) are ignored.
2929 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2930 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2932 PH/29 [Removed. This was a change that I later backed out, and forgot to
2933 correct the ChangeLog entry (that I had efficiently created) before
2934 committing the later change.]
2936 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2937 address errors (4xx response to RCPT) over the same connection as other
2938 messages routed to the same hosts. Such deliveries are always "forced",
2939 so retry times are not inspected. This resulted in far too many retries
2940 for the affected addresses. The effect occurred only when there were more
2941 hosts than the hosts_max_try setting in the smtp transport when it had
2942 the 4xx errors. Those hosts that it had tried were not added to the list
2943 of hosts for which the message was waiting, so if all were tried, there
2944 was no problem. Two fixes have been applied:
2946 (i) If there are any address or message errors in an SMTP delivery, none
2947 of the hosts (tried or untried) are now added to the list of hosts
2948 for which the message is waiting, so the message should not be a
2949 candidate for sending over the same connection that was used for a
2950 successful delivery of some other message. This seems entirely
2951 reasonable: after all the message is NOT "waiting for some host".
2952 This is so "obvious" that I'm not sure why it wasn't done
2953 previously. Hope I haven't missed anything, but it can't do any
2954 harm, as the worst effect is to miss an optimization.
2956 (ii) If, despite (i), such a delivery is accidentally attempted, the
2957 routing retry time is respected, so at least it doesn't keep
2958 hammering the server.
2960 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2961 in ${readsocket because some servers need this prod.
2963 PH/32 Added some extra debug output when updating a wait-xxx database.
2965 PH/33 The hint "could be header name not terminated by colon", which has been
2966 given for certain expansion errors for a long time, was not being given
2967 for the ${if def:h_colon_omitted{... case.
2969 PH/34 The spec says: "With one important exception, whenever a domain list is
2970 being scanned, $domain contains the subject domain." There was at least
2971 one case where this was not true.
2973 PH/35 The error "getsockname() failed: connection reset by peer" was being
2974 written to the panic log as well as the main log, but it isn't really
2975 panic-worthy as it just means the connection died rather early on. I have
2976 removed the panic log writing for the ECONNRESET error when getsockname()
2979 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2980 runs only) independently of the message's sender address. This meant
2981 that, if the 4xx error was in fact related to the sender, a different
2982 message to the same recipient with a different sender could confuse
2983 things. In particular, this can happen when sending to a greylisting
2984 server, but other circumstances could also provoke similar problems.
2985 I have changed the default so that the retry time for these errors is now
2986 based a combination of the sender and recipient addresses. This change
2987 can be overridden by setting address_retry_include_sender=false in the
2990 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2991 remote server are returned as part of bounce messages. This was not
2992 happening for LMTP over a pipe (the lmtp transport), but now it is the
2993 same for both kinds of LMTP.
2995 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2996 in header lines that were in fact CNAMEs. This is no longer the case.
2998 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2999 and queue runs started by the daemon processed all messages. This has
3000 been fixed so that -R and -S can now usefully be given with -q<time>.
3002 PH/40 Import PCRE release 6.7 (fixes some bugs).
3004 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3006 PH/42 Give an error if -q is specified more than once.
3008 PH/43 Renamed the variables $interface_address and $interface_port as
3009 $received_ip_address and $received_port, to make it clear that these
3010 values apply to message reception, and not to the outgoing interface when
3011 a message is delivered. (The old names remain recognized, of course.)
3013 PH/44 There was no timeout on the connect() call when using a Unix domain
3014 socket in the ${readsocket expansion. There now is.
3016 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3017 be meaningful with "accept".
3019 SC/01 Eximstats V1.43
3020 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3022 SC/02 Eximstats V1.44
3023 Use a glob alias rather than an array ref in the generated
3024 parser. This improves both readability and performance.
3026 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3027 Collect SpamAssassin and rejection statistics.
3028 Don't display local sender or destination tables unless
3029 there is data to show.
3030 Added average volumes into the top table text output.
3032 SC/04 Eximstats V1.46
3033 Collect data on the number of addresses (recipients)
3034 as well as the number of messages.
3036 SC/05 Eximstats V1.47
3037 Added 'Message too big' to the list of mail rejection
3038 reasons (thanks to Marco Gaiarin).
3040 SC/06 Eximstats V1.48
3041 Mainlog lines which have GMT offsets and are too short to
3042 have a flag are now skipped.
3044 SC/07 Eximstats V1.49 (Alain Williams)
3045 Added the -emptyok flag.
3047 SC/08 Eximstats V1.50
3048 Fixes for obtaining the IP address from reject messages.
3050 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3051 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3052 whitespace changes from 4.64-PH/27
3054 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3057 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3058 are found, allow negative numbers in numeric criteria)
3060 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3062 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3063 to match changes made in 4.64-PH/43
3065 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3067 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3068 "domain" is actually an IP address, and force a failure. This locks out
3069 those revolvers/nameservers that support "A-for-A" lookups, in
3070 contravention of the specifications.
3072 PH/48 When a host name was looked up from an IP address, and the subsequent
3073 forward lookup of the name timed out, the host name was left in
3074 $sender_host_name, contrary to the specification.
3076 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3077 restricted to single-key lookups, Exim was not diagnosing an error if
3078 * or *@ was used with a query-style lookup.
3080 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3082 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3083 long ago, but noone interested enough thought of it. Let's just say that
3084 the "1.1" means that there are some new functions that weren't there at
3085 some point in the past.
3087 PH/51 Error processing for expansion failure of helo_data from an smtp
3088 transport during callout processing was broken.
3090 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3091 tested/used via the -bh/-bhc/-bs options.
3093 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3094 bug, fixed in subsequent PCRE releases).
3096 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3097 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3099 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3105 SC/01 Use a glob alias rather than an array ref in eximstats generated
3106 parser. This improves both readability and performance.
3108 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3109 Don't display local sender or destination tables in eximstats unless
3110 there is data to show.
3111 Added average volumes into the eximstats top table text output.
3113 SC/03 Collect data on the number of addresses (recipients) as well
3114 as the number of messages in eximstats.
3116 TF/01 Correct an error in the documentation for the redirect router. Exim
3117 does (usually) call initgroups() when daemonizing.
3119 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3120 with consistent privilege compared to when running as a daemon.
3122 TF/03 Note in the spec that $authenticated_id is not set for local
3123 submissions from trusted users.
3125 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3126 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3128 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3129 by adding some example configuration directives to the default
3130 configuration file. A little bit of work is required to uncomment the
3131 directives and define how usernames and passwords are checked, but
3132 there is now a framework to start from.
3134 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3135 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3136 without this. I don't know how relevant this is to other LDAP libraries.
3138 PH/02 Add the verb name to the "unknown ACL verb" error.
3140 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3142 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3144 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3145 directory not expanded when it should be if an expanded home directory
3146 was set for the address (which is overridden by the transport).
3148 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3151 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3152 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3153 because it is too late at that time, and has no effect.
3155 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3156 security issue with \' (bugzilla #107). I could not use the
3157 PQescapeStringConn() function, because it needs a PGconn value as one of
3160 PH/08 When testing addresses using -bt, indicate those final addresses that
3161 are duplicates that would not cause an additional delivery. At least one
3162 person was confused, thinking that -bt output corresponded to deliveries.
3163 (Suppressing duplicates isn't a good idea as you lose the information
3164 about possibly different redirections that led to the duplicates.)
3166 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3167 systems where poll() doesn't work, in particular OS X.
3169 PH/10 Added more information to debugging output for retry time not reached.
3171 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3172 operations in malware.c.
3174 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3177 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3178 syslog_duplication set false, log lines that would normally be written
3179 both the the main log and to the reject log were not written to syslog at
3182 PH/14 In the default configuration, change the use of "message" in ACL warn
3183 statements to "add_header".
3185 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3186 not followed by a command (e.g. "seen endif").
3188 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3189 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3192 PH/17 Added extra conditions to the default value of delay_warning_condition
3196 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3197 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3198 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3201 The Auto-Submitted: and various List- headers are standardised, whereas I
3202 don't think Precedence: ever was.
3204 PH/18 Refactored debugging code in route_finduser() to show more information,
3205 in particular, the error code if getpwnam() issues one.
3207 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3208 This is apparently needed in addition to the PH/07 change above to avoid
3209 any possible encoding problems.
3211 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3212 but not after initializing Perl.
3214 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3215 output them only if debugging. By default they are written stderr,
3216 apparently, which is not desirable.
3218 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3221 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3224 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3226 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3227 authenticated or an ident call has been made. Suppress the default
3228 values for $authenticated_id and $authenticated_sender (but permit -oMai
3229 and -oMas) when testing with -bh.
3231 PH/24 Re-jigged the order of the tests in the default configuration so that the
3232 tests for valid domains and recipients precede the DNS black list and CSA
3233 tests, on the grounds that those ones are more expensive.
3235 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3236 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3237 This bug exists in every version of Exim that I still have, right back to
3240 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3241 However, an attempt to turn on case-sensitivity in a regex key by
3242 including (?-i) didn't work because the subject string was already
3243 lowercased, and the effects were non-intuitive. It turns out that a
3244 one-line patch can be used to allow (?-i) to work as expected.
3250 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3251 other effects) broke the use of negated acl sub-conditions.
3253 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3256 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3257 "Deny" causes Exim to reject the incoming connection with a 554 error.
3258 Unfortunately, if there is a major crisis, such as a disk failure,
3259 tcp-wrappers gives "deny", whereas what one would like would be some
3260 kind of temporary error. A kludge has been added to help with this.
3261 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3262 554 error is used if errno is still zero or contains ENOENT (which occurs
3263 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3266 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3268 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3269 errors. Otherwise a message that provokes a temporary error (when other
3270 messages do not) can cause a whole host to time out.
3272 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3273 addresses were routed directly to files or pipes from a redirect router.
3274 File deliveries just didn't batch; pipe deliveries might have suffered
3277 PH/06 A failure to get a lock for a hints database would erroneously always say
3278 "Failed to get write lock", even when it was really a read lock.
3280 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3281 of 0600. This has been changed to use the value of the lockfile_mode
3282 option (which defaults to 0600).
3284 PH/08 Applied small patch from the Sieve maintainer.
3286 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3287 folder from quota calculations, a direct delivery into this folder messed
3288 up the contents of the maildirsize file. This was because the regex was
3289 used only to exclude .Trash (or whatever) when the size of the mailbox
3290 was calculated. There was no check that a delivery was happening into an
3291 excluded directory. This bug has been fixed by ignoring all quota
3292 processing for deliveries into excluded directories.
3294 PH/10 Added the maildirfolder_create_regex option to appendfile.
3300 PH/01 The code for finding all the local interface addresses on a FreeBSD
3301 system running IPv6 was broken. This may well have applied to all BSD
3302 systems, as well as to others that have similar system calls. The broken
3303 code found IPv4 interfaces correctly, but gave incorrect values for the
3304 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3305 that it would not match correctly against @[] and not recognize the IPv6
3308 PH/02 The ipliteral router was not recognizing addresses of the form user@
3309 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3311 PH/03 Added disable_ipv6.
3313 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3314 decoded form, because it is most often used to construct To: headers
3315 lines in autoreplies, and the decoded form may well be syntactically
3316 invalid. However, $reply_address has leading white space removed, and all
3317 newlines turned into spaces so that the autoreply transport does not
3320 PH/05 If group was specified without a user on a router, and no group or user
3321 was specified on a transport, the group from the router was ignored.
3323 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3324 for visible compile-time settings that can be used to change these
3325 numbers, for those that want even more. Backwards compatibility with old
3326 spool files has been maintained. However, going back to a previous Exim
3327 release will lost any variables that are in spool files.
3329 PH/07 Two small changes when running in the test harness: increase delay when
3330 passing a TCP/IP connection to a new process, in case the original
3331 process has to generate a bounce, and remove special handling of
3332 127.0.0.2 (sic), which is no longer necessary.
3334 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3335 be the same on different OS.
3337 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3340 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3341 whether --show-vars was specified or not
3343 JJ/02 exipick: Added support for new ACL variable spool format introduced
3346 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3347 syntactically invalid From: or Reply-to: line, and a filter used this to
3348 generate an autoreply, and therefore failed to obtain an address for the
3349 autoreply, Exim could try to deliver to a non-existent relative file
3350 name, causing unrelated and misleading errors. What now happens is that
3351 it logs this as a hard delivery error, but does not attempt to create a
3354 PH/11 The exinext utility has a -C option for testing purposes, but although
3355 the given file was scanned by exinext itself; it wasn't being passed on
3356 when Exim was called.
3358 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3359 an end-of-file indication when reading a command response.
3361 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3362 compiled. In many other places in Exim, IPv6 addresses are always
3363 recognized, so I have changed this. It also means that IPv4 domain
3364 literals of the form [IPV4:n.n.n.n] are now always recognized.
3366 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3367 used if the router is not running as root, for example, when verifying at
3368 ACL time, or when using -bh. The debugging output from this situation was
3369 non-existent - all you got was a failure to exec. I have made two
3372 (a) Failures to set uid/gid, the current directory, or a process leader
3373 in a subprocess such as that created by queryprogram now generate
3374 suitable debugging output when -d is set.
3376 (b) The queryprogram router detects when it is not running as root,
3377 outputs suitable debugging information if -d is set, and then runs
3378 the subprocess without attempting to change uid/gid.
3380 PH/15 Minor change to Makefile for building test_host (undocumented testing
3383 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3384 additional section of a DNS packet that returns MX or SRV records.
3385 Instead, it always explicitly searches for A/AAAA records. This avoids
3386 major problems that occur when a DNS server includes only records of one
3387 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3388 fixed another bug: if SRV records were looked up and the corresponding
3389 address records were *not* found in the additional section, the port
3390 values from the SRV records were lost.
3392 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3393 using the correct key (the original address) when searching the retry
3394 rules in order to find which one to use for generating the retry hint.
3396 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3397 adding the default one. Similarly, if it contains a Reply-To: header, the
3398 errors_reply_to option, if set, is not used.
3400 PH/19 When calculating a retry time, Exim used to measure the "time since
3401 failure" by looking at the "first failed" field in the retry record. Now
3402 it does not use this if it is later than than the arrival time of the
3403 message. Instead it uses the arrival time. This makes for better
3404 behaviour in cases where some deliveries succeed, thus re-setting the
3405 "first failed" field. An example is a quota failure for a huge message
3406 when small messages continue to be delivered. Without this change, the
3407 "time since failure" will always be short, possible causing more frequent
3408 delivery attempts for the huge message than are intended.
3409 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3411 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3412 $1, $2, $3) because the numerical variables can be reset during some
3413 expansion items (e.g. "match"), thereby losing the authentication data.
3415 PH/21 Make -bV show the size of off_t variables so that the test suite can
3416 decide whether to run tests for quotas > 2G.
3418 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3419 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3420 filecount value is greater than 2G or if a quota value is greater than 2G
3421 on a system where the size of off_t is not greater than 4, a panic error
3424 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3425 never match. The debug and -bh output now contains an explicit error
3426 message indicating a malformed IPv4 address or mask.
3428 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3429 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3430 PH/23 above applies.
3432 PH/25 Do not write to syslog when running in the test harness. The only
3433 occasion when this arises is a failure to open the main or panic logs
3434 (for which there is an explicit test).
3436 PH/26 Added the /no_tell option to "control=freeze".
3438 PH/27 If a host name lookup failed very early in a connection, for example, if
3439 the IP address matched host_lookup and the reverse lookup yielded a name
3440 that did not have a forward lookup, an error message of the form "no IP
3441 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3442 could be logged. Now it outputs the IP address instead of "NULL".
3444 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3445 allows the sender and the authenticated sender to be set when
3446 submitting a message from within Exim. Since child_open_exim() is
3447 documented for local_scan(), the new function should be too.
3449 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3450 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3451 results in an empty string is now treated as unset.
3453 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3455 PH/31 Added sender_verify_fail logging option.
3457 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3458 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3459 tidied the source and removed it altogether.
3461 PH/33 When a queue run was abandoned because the load average was too high, a
3462 log line was always written; now it is written only if the queue_run log
3463 selector is set. In addition, the log line for abandonment now contains
3464 information about the queue run such as the pid. This is always present
3465 in "start" and "stop" lines but was omitted from the "abandon" line.
3467 PH/34 Omit spaces between a header name and the colon in the error message that
3468 is given when verify = headers_syntax fails (if there are lots of them,
3469 the message gets confusing).
3471 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3472 names, as there are now some PTR records that contain slashes. This check
3473 is only to protect against broken name servers that fall over on strange
3474 characters, so the fact that it applies to all lookups doesn't matter.
3476 PH/36 Now that the new test suite is complete, we can remove some of the
3477 special code in Exim that was needed for the old test suite. For example,
3478 sorting DNS records because real resolvers return them in an arbitrary
3479 order. The new test suite's fake resolver always returns records in the
3482 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3483 bounces) except when queue_only is set, to avoid logging races between
3484 the different processes.
3486 PH/38 Panic-die if .include specifies a non-absolute path.
3488 PH/39 A tweak to the "H" retry rule from its user.
3490 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3491 a label. They prevented compilation on older perls.
3493 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3494 a warning to be raised on newish perls.
3496 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3497 on queue. Changes to match documented behaviour of showing count of
3498 messages matching specified criteria.
3500 PH/40 Changed the default ident timeout from 30s to 5s.
3502 PH/41 Added support for the use of login_cap features, on those BSD systems
3503 that have them, for controlling the resources used by pipe deliveries.
3505 PH/42 The content-scanning code uses fopen() to create files in which to put
3506 message data. Previously it was not paying any attention to the mode of
3507 the files. Exim runs with umask(0) because the rest of the code creates
3508 files with open(), and sets the required mode explicitly. Thus, these
3509 files were ending up world-writeable. This was not a big issue, because,
3510 being within the spool directory, they were not world-accessible. I have
3511 created a function called modefopen, which takes an additional mode
3512 argument. It sets umask(777), creates the file, chmods it to the required
3513 mode, then resets the umask. All the relevant calls to fopen() in the
3514 content scanning code have been changed to use this function.
3516 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3517 to 24 hours. This avoids potential overflow problems when processing G
3518 and H retry rules. I suspect nobody ever tinkers with this value.
3520 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3522 PH/45 When the plaintext authenticator is running as a client, the server's
3523 challenges are checked to ensure they are valid base64 strings. By
3524 default, the authentication attempt is cancelled if an invalid string is
3525 received. Setting client_ignore_invalid_base64 true ignores these errors.
3526 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3527 they are received. Thus, the responses can be made to depend on the
3528 challenges. If an invalid string is ignored, an empty string is placed in
3531 PH/46 Messages that are created by the autoreply transport now contains a
3532 References: header, in accordance with RFCs 2822 and 3834.
3534 PH/47 Added authenticated_sender_force to the smtp transport.
3536 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3538 PH/49 Installed latest patch from the Sieve maintainer.
3540 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3541 was also set, the appendfile transport was unnecessarily scanning a
3542 directory of message files (e.g. for maildir delivery) to find the count
3543 of files (along with the size), even though it did not need this
3544 information. It now does the scan only if it needs to find either the
3545 size of the count of files.
3547 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3549 PH/52 Two bugs concerned with error handling when the smtp transport is
3552 (i) Exim was not creating retry information for temporary errors given
3553 for individual recipients after the DATA command when the smtp transport
3554 was used in LMTP mode. This meant that they could be retried too
3555 frequently, and not timed out correctly.
3557 (ii) Exim was setting the flag that allows error details to be returned
3558 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3559 recipients that were returned after the DATA command.
3561 PH/53 This is related to PH/52, but is more general: for any failing address,
3562 when detailed error information was permitted to be returned to the
3563 sender, but the error was temporary, then after the final timeout, only
3564 "retry timeout exceeded" was returned. Now it returns the full error as
3565 well as "retry timeout exceeded".
3567 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3568 do this, and (what is worse) MTAs that accept it.
3570 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3571 will now be deprecated.
3573 PH/56 New os.c-cygwin from the Cygwin maintainer.
3575 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3576 formats (previously only available in exim formats via -bpr, -bpru,
3577 and -bpra. Now also available in native and exiqgrep formats)
3579 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3580 with very large, slow to parse queues
3582 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3584 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3586 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3587 responses to authentication challenges, though it was showing the
3588 challenges; (ii) I've removed the CR characters from the debug output for
3591 PH/58 Allow for the insertion of a newline as well as a space when a string
3592 is turned into more than one encoded-word during RFC 2047 encoding. The
3593 Sieve code now uses this.
3595 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3596 data_4xx, lost_connection, tls_required.
3598 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3599 message was being sent as an SMTP response.
3601 PH/61 Add -l and -k options to exicyclog.
3603 PH/62 When verifying, if an address was redirected to one new address, so that
3604 verification continued, and the new address failed or deferred after
3605 having set something in $address_data, the value of $address_data was not
3606 passed back to the ACL. This was different to the case when no
3607 redirection occurred. The value is now passed back in both cases.
3609 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3610 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3611 use login_cap.h, so on its own it isn't the distinguishing feature. The
3612 new name refers directly to the setclassresources() function.
3614 PH/65 Added configuration files for NetBSD3.
3616 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3618 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3619 is preferred over IPv4.
3621 PH/68 The bounce_return_message and bounce_return_body options were not being
3622 honoured for bounces generated during the reception of non-SMTP messages.
3623 In particular, this applied to messages rejected by the ACL. This bug has
3624 been fixed. However, if bounce_return_message is true and bounce_return_
3625 body is false, the headers that are returned for a non-SMTP message
3626 include only those that have been read before the error was detected.
3627 (In the case of an ACL rejection, they have all been read.)
3629 PH/69 The HTML version of the specification is now built in a directory called
3630 spec_html instead of spec.html, because the latter looks like a path with
3631 a MIME-type, and this confuses some software.
3633 PH/70 Catch two compiler warnings in sieve.c.
3635 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3636 function verify_get_ident() calls ip_connect() to connect a socket, but
3637 if the "connect()" function timed out, ip_connect() used to close the
3638 socket. However, verify_get_ident() also closes the socket later, and in
3639 between Exim writes to the log, which may get opened at this point. When
3640 the socket was closed in ip_connect(), the log could get the same file
3641 descriptor number as the socket. This naturally causes chaos. The fix is
3642 not to close the socket in ip_connect(); the socket should be closed by
3643 the function that creates it. There was only one place in the code where
3644 this was missing, in the iplookup router, which I don't think anybody now
3645 uses, but I've fixed it anyway.
3647 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3648 well as to direct DNS lookups. Otherwise the handling of names in host
3649 lists is inconsistent and therefore confusing.
3655 PH/01 Two changes to the default runtime configuration:
3657 (1) Move the checks for relay_from_hosts and authenticated clients from
3658 after to before the (commented out) DNS black list checks.
3660 (2) Add control=submission to the relay_from_hosts and authenticated
3661 clients checks, on the grounds that messages accepted by these
3662 statements are most likely to be submissions.
3664 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3666 (1) Generate an error if the third argument for the ${prvs expansion is
3669 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3672 (3) Reset the variables that are obtained from the first argument of
3673 ${prvscheck and used in the second argument before leaving the code,
3674 because their memory is reclaimed, so using them afterwards may do
3677 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3678 one (it's much easier than Tom thought :-).
3680 (5) Because of (4), we can now allow for the use of $prvscheck_result
3681 inside the third argument.
3683 PH/03 For some reason, the default setting of PATH when running a command from
3684 a pipe transport was just "/usr/bin". I have changed it to
3687 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3688 anything to be listed in the output from -bV.
3690 PH/05 When a filter generated an autoreply, the entire To: header line was
3691 quoted in the delivery log line, like this:
3693 => >A.N.Other <ano@some.domain> <original@ddress> ...
3695 This has been changed so that it extracts the operative address. There
3696 may be more than one such address. If so, they are comma-separated, like
3699 => >ano@some.domain,ona@other.domain <original@ddress> ...
3701 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3702 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3703 not being looked up in the rDNS to get a host name, Exim was showing the
3704 IP address twice in Received: lines, even though the IP addresses were
3705 identical. For example:
3707 Received: from [1.2.3.4] (helo=[1.2.3.4])
3709 However, if the real host name was known, it was omitting the HELO data
3710 if it matched the actual IP address. This has been tidied up so that it
3711 doesn't show the same IP address twice.
3713 PH/07 When both +timestamp and +memory debugging was on, the value given by
3714 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3715 called by the debug printing, thereby overwriting the timestamp buffer.
3716 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3718 PH/08 When the original message was included in an autoreply transport, it
3719 always said "this is a copy of the message, including all the headers",
3720 even if body_only or headers_only was set. It now gives an appropriate
3723 PH/09 Applied a patch from the Sieve maintainer which:
3725 o fixes some comments
3726 o adds the (disabled) notify extension core
3727 o adds some debug output for the result of if/elsif tests
3728 o points to the current vacation draft in the documentation
3729 and documents the missing references header update
3733 o fixes a bug in processing the envelope test (when testing
3734 multiple envelope elements, the last element determined the
3737 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3738 Electronic Mail") by including:
3740 Auto-submitted: auto-generated
3742 in the messages that it generates (bounce messages and others, such as
3743 warnings). In the case of bounce messages for non-SMTP messages, there was
3744 also a typo: it was using "Auto_submitted" (underscore instead of
3745 hyphen). Since every message generated by Exim is necessarily in response
3746 to another message, thes have all been changed to:
3748 Auto-Submitted: auto-replied
3750 in accordance with these statements in the RFC:
3752 The auto-replied keyword:
3754 - SHOULD be used on messages sent in direct response to another
3755 message by an automatic process,
3757 - MUST NOT be used on manually-generated messages,
3759 - MAY be used on Delivery Status Notifications (DSNs) and Message
3760 Disposition Notifications (MDNs),
3762 - MUST NOT be used on messages generated by automatic or periodic
3763 processes, except for messages which are automatic responses to
3766 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3767 to the default Received: header definition.
3769 PH/12 Added log selector acl_warn_skipped (default on).
3771 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3772 variables because (a) they are in the wrong storage pool and (b) even if
3773 they were copied, it wouldn't work properly because of the caching.
3775 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3776 checking when decoding. Apparently there are clients that generate
3777 overlong encoded strings. Why am I not surprised?
3779 PH/15 If the first argument of "${if match_address" was not empty, but did not
3780 contain an "@" character, Exim crashed. Now it writes a panic log message
3781 and treats the condition as false.
3783 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3785 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3786 extension "envelope-auth". The code is finished and in agreement with
3787 other implementations, but there is no documentation so far and in fact,
3788 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3789 not changing the active code.
3791 Print executed "if" and "elsif" statements when debugging is used. This
3792 helps a great deal to understand what a filter does.
3794 Document more things not specified clearly in RFC3028. I had all this
3795 sorted out, when out of a sudden new issues came to my mind. Oops."
3797 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3800 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3801 canonical form (as documented). However, after a host name lookup from
3802 the IP address, check_host() was doing a simple string comparison with
3803 addresses acquired from the DNS when checking that the found name did
3804 have the original IP as one of its addresses. Since any found IPv6
3805 addresses are likely to be in abbreviated form, the comparison could
3806 fail. Luckily, there already exists a function for doing the comparison
3807 by converting both addresses to binary, so now that is used instead of
3808 the text comparison.
3810 PH/20 There was another similar case to PH/19, when a complete host name was
3811 given in a host list; looking up its IP address could give an abbreviated
3812 form, whereas the current host's name might or might not be abbreviated.
3813 The same fix has been applied.
3819 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3820 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3823 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3824 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3826 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3828 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3829 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3830 Also, for "Auto-Submitted", treat anything other than "no" as
3831 non-personal, in accordance with RFC 3834. (Previously it treated
3832 anything starting "auto-" as non-personal.)
3834 TF/01 The control=submission/name=... option had a problem with syntax
3835 errors if the name included a slash character. The /name= option
3836 now slurps the rest of the string, so it can include any characters
3837 but it must come last in the list of options (after /sender_retain
3840 PH/05 Some modifications to the interface to the fake nameserver for the new
3848 TK/01 Added the "success_on_redirect" address verification option. See
3849 NewStuff for rationale and an example.
3851 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3853 PH/02 Patch to exigrep to allow it to work on syslog lines.
3855 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3856 fread() to skip over the body file's header line, because in Cygwin the
3857 header line is locked and is inaccessible.
3859 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3860 co-exist for some time) to make it clear that it is the Exim ID that is
3861 referenced, not the Message-ID: header line.
3863 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3864 string_format() function, because snprintf() does not exist on all
3867 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3868 ${stat: expansion item.
3870 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3871 protocol synchronization error", to keep the pedants happy.
3873 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3874 well as for IRIX systems, when gcc is being used. See the host.c source
3877 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3879 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3882 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3883 transport and to the smtp transport in LMTP mode.
3885 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3887 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3888 run a filter in a subprocess. This could lead to confusion in subsequent
3889 lookups in the parent process. There should also be a search_tidyup() at
3890 the end of the subprocess.
3892 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3893 only if the host matched helo_try_verify_hosts, which caused the
3894 verification to occur when the EHLO/HELO command was issued. The ACL just
3895 tested the remembered result. Now, if a previous verification attempt has
3896 not happened, "verify = helo" does it there and then.
3898 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3900 TK/03 Fix log output including CR from clamd.
3902 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3903 exist provoked a memory error which could cause a segfault.
3905 PH/15 Installed PCRE 6.2
3907 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3909 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3910 of the problem. Specifically, suggested +O2 rather than +O1 for the
3913 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3915 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3916 record, the port was not correctly logged when the outgoing_port log
3917 selector was set (it logged the transort's default port).
3919 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3920 fallback_hosts, and "hosts" in the smtp transport.
3922 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3923 host errors such as "Connection refused".
3925 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3926 authentication with radiusclient 0.4.9:
3928 - Error returned from rc_read_config was caught wrongly
3929 - Username/password not passed on to radius server due to wrong length.
3931 The presumption is that some radiusclient API changes for 4.51/PH/17
3932 were not taken care of correctly. The code is still untested by me (my
3933 Linux distribution still has 0.3.2 of radiusclient), but it was
3934 contributed by a Radius user.
3936 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3937 expanding the "port" option of the smtp transport.
3939 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3940 while reading a MIME header. Thanks to Tom Hughes for a patch.
3942 PH/24 Include config.h inside local_scan.h so that configuration settings are
3945 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3946 that in an ACL for RCPT (for example), you can examine exactly what was
3949 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3950 commands, but it was not correctly comparing the address with the actual
3951 client host address. Thus, it would show the EHLO address in Received:
3952 header lines when this was not necessary.
3954 PH/27 Added the % operator to ${eval:}.
3956 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3957 it should be ignoring failures (because with -C, for example, it has lost
3958 privilege). It wasn't ignoring creation failures other than "already
3961 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3964 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3965 because an input line was too long, either on its own, or by virtue of
3966 too many continuations, the temporary file was not being removed, and the
3967 return code was incorrect.
3969 PH/31 Missing "BOOL" in function definition in filtertest.c.
3971 PH/32 Applied Sieve patches from the maintainer.
3973 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3975 PH/33 Added "verify = not_blind".
3977 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3978 Local/Makefile (with some defaults set). These are used in built scripts
3979 such as exicyclog, but they have never been used in the exim_install
3980 script (though there are many overriding facilities there). I have
3981 arranged that the exim_install script now takes note of these two
3984 PH/35 Installed configuration files for Dragonfly.
3986 PH/36 When a locally submitted message by a trusted user did not contain a
3987 From: header, and the sender address was obtained from -f or from an SMTP
3988 MAIL command, and the trusted user did not use -F to supply a sender
3989 name, $originator_name was incorrectly used when constructing a From:
3990 header. Furthermore, $originator_name was used for submission mode
3991 messages from external hosts without From: headers in a similar way,
3992 which is clearly wrong.
3994 PH/37 Added control=suppress_local_fixups.
3996 PH/38 When log_selector = +received_sender was set, and the addition of the
3997 sender made the log line's construction buffer exactly full, or one byte
3998 less than full, an overflow happened when the terminating "\n" was
4001 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4002 when the result of a list match is failure because a DNS lookup failed.
4004 PH/40 RM_COMMAND is now used in the building process.
4006 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4007 the "build-* directories that it finds.
4009 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4010 address in a domain literal was a prefix of an interface address.
4012 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4013 when verifying a sender address, unless rewrite_headers is false.
4015 PH/44 Wrote a long comment about why errors_to addresses are verified as
4016 recipients, not senders.
4018 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4019 the ratelimit ACL was added.
4021 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4023 PH/46 Added extra information about PostgreSQL errors to the error string.
4025 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4026 avoiding the need to install special zones in a real server. This is
4027 backwards compatible; if it can't find the fake resolver, it drops back.
4028 Thus, both old and new test suites can be run.
4030 TF/02 Added util/ratelimit.pl
4032 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4033 clock is set back in time.
4035 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4036 Candler <B.Candler@pobox.com>.
4038 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4039 OK for addresses that are the result of redirections.
4041 PH/48 A number of further additions for the benefit of the new test suite,
4042 including a fake gethostbyname() that interfaces to the fake DNS resolver
4045 TF/06 The fix for widen_domains has also been applied to qualify_single and
4046 search_parents which are the other dnslookup options that can cause
4049 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4052 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4054 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4055 service" if it encountered an MX record with an empty target hostname.
4056 The message is now "an MX or SRV record indicated no SMTP service".
4058 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4059 if the DNS of the sending site is misconfigured. This is quite a
4060 common situation. This change restores the behaviour of treating a
4061 helo verification defer as a failure.
4063 PH/51 If self=fail was set on a router, the bounce message did not include the
4064 actual error message.
4070 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4072 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4073 command itself ended in error, the underlying message about the transport
4074 filter timeout was being overwritten with the pipe command error. Now the
4075 underlying error message should be appended to the second error message.
4077 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4079 PH/02 Reduce the amount of output that "make" produces by default. Full output
4080 can still be requested.
4082 PH/03 The warning log line about a condition test deferring for a "warn" verb
4083 was being output only once per connection, rather than after each
4084 occurrence (because it was using the same function as for successful
4085 "warn" verbs). This seems wrong, so I have changed it.
4087 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4088 it should not have, which might have caused a crash in the right
4089 circumstances, but probably never did.
4091 PH/04 Installed a modified version of Tony Finch's patch to make submission
4092 mode fix the return path as well as the Sender: header line, and to
4093 add a /name= option so that you can make the user's friendly name appear
4096 TF/03 Added the control = fakedefer ACL modifier.
4098 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4099 Mark Lowes for thorough testing.
4101 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4103 TK/03 Merged latest SRS patch from Miles Wilton.
4105 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4106 with the definition in sysexits.h (which is #included earlier).
4107 Fortunately, Exim does not actually use EX_OK. The code used to try to
4108 preserve the sysexits.h value, by assuming that macro definitions were
4109 scanned for macro replacements. I have been disabused of this notion,
4110 so now the code just undefines EX_OK before #including unistd.h.
4112 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4113 in the smtp transport. When a block could not be written in a single
4114 write() function, the timeout was being re-applied to each part-write.
4115 This seems wrong - if the receiver was accepting one byte at a time it
4116 would take for ever. The timeout is now adjusted when this happens. It
4117 doesn't have to be particularly precise.
4119 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4120 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4122 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4123 without a domain if the check to <postmaster@domain> fails.
4125 SC/01 Eximstats: added -xls and the ability to specify output files
4126 (patch written by Frank Heydlauf).
4128 SC/02 Eximstats: use FileHandles for outputting results.
4130 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4132 SC/04 Eximstats: fixed display of large numbers with -nvr option
4134 SC/05 Eximstats: fixed merging of reports with empty tables.
4136 SC/06 Eximstats: added the -include_original_destination flag
4138 SC/07 Eximstats: removed tabs and trailing whitespace.
4140 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4142 TK/06 MBOX spool code: Add real "From " MBOX separator line
4143 so the .eml file is really in mbox format (even though
4144 most programs do not really care). Patch from Alex Miller.
4146 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4147 The latter is generated from $received_to and is only set if the
4148 message has one envelope recipient. SA can use these headers,
4149 obviously out-of-the-box. Patch from Alex Miller.
4151 PH/08 The ${def test on a variable was returning false if the variable's
4152 value was "0", contrary to what the specification has always said!
4153 The result should be true unless the variable is empty.
4155 PH/09 The syntax error of a character other than { following "${if
4156 def:variable_name" (after optional whitespace) was not being diagnosed.
4157 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4158 accidental colon was present, for example, could give incorrect results.
4160 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4161 result is used (not including appendfile, where other changes are about
4164 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4165 This involved changing a lot of size variables from int to off_t. It
4166 should work with maildirs and everything.
4168 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4169 spamd dying while we are connected to it.
4171 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4174 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4175 function for BDB 4.3.
4177 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4179 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4180 the sources for PCRE have been re-arranged and more files are now
4183 PH/15 The code I had for printing potentially long long variables in PH/11
4184 above was not the best (it lost precision). The length of off_t variables
4185 is now inspected at build time, and an appropriate printing format (%ld
4186 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4187 to be "long long int" or "long int". This is needed for the internal
4188 formatting function string_vformat().
4190 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4191 the configuration file to be ":syslog", then the script "guesses" where
4192 the logs files are, rather than using the compiled in default. In our
4193 case the guess is not the same as the compiled default, so the script
4194 suddenly stopped working when I started to use syslog. The patch checks
4195 to see if log_file_path is "". If so, it attempts to read it from exim
4196 with no configuration file to get the compiled in version, before it
4197 falls back to the previous guessing code."
4199 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4200 implementing BATV in an Exim configuration. See NewStuff for the gory
4203 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4204 Makefile that are specific to HP-UX.
4206 PH/18 If the "use_postmaster" option was set for a recipient callout together
4207 with the "random" option, the postmaster address was used as the MAIL
4208 FROM address for the random test, but not for the subsequent recipient
4209 test. It is now used for both.
4211 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4212 patch removes a few documentation additions to RFC 3028, because the
4213 latest draft now contains them. It adds the new en;ascii-case comparator
4214 and a new error check for 8bit text in MIME parts. Comparator and
4215 require names are now matched exactly. I enabled the subaddress
4216 extension, but it is not well tested yet (read: it works for me)."
4218 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4219 rework some of the code of TK/09 above to avoid the hardwired use of
4220 "%lld" and "long long". Replaced the call to snprintf() with a call to
4223 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4224 records point to non-existent hosts", "retry timeout exceeded", and
4225 "retry time not reached for any host after a long failure period".
4227 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4228 experimental DomainKeys support:
4230 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4231 (2) On an error such as an illegally used "control", the wrong name for
4232 the control was given.
4234 These problems did NOT occur unless DomainKeys support was compiled.
4236 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4238 PH/24 Added ${if match_ip condition.
4240 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4241 databases so that it will be absolutely obvious if a crash occurs in the
4242 DB library. This is a regular occurrence (often caused by mis-matched
4245 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4246 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4247 picked up on a user's system that detects such things. There doesn't seem
4248 to be a gcc warning option for this - only an attribute that has to be
4249 put on the function's prototype. It seems that in Fedora Core 4 they have
4250 set this on a number of new functions. No doubt there will be more in due
4253 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4254 specify a transport. However, if an address that was verified by such a
4255 router was the subject of a callout, Exim crashed because it tried to
4256 read the rcpt_include_affixes from the non-existent transport. Now it
4257 just assumes that the setting of that option is false. This bug was
4258 introduced by 4.51/PH/31.
4260 PH/28 Changed -d+all to exclude +memory, because that information is very
4261 rarely of interest, but it makes the output a lot bigger. People tend to
4262 do -d+all out of habit.
4264 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4265 code in os-type was giving problems when libc.so lives in lib64, like on
4268 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4269 aren't the modern standard, and it seems that some systems' include files
4270 don't always have them. Exim was already checking for some of the newer
4271 ones like T_AAAA, and defining it itself. I've added checks for all the
4272 record types that Exim uses.
4274 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4275 not automatically generating a new one, as it is supposed to. This
4276 prevented TLS from working. If the file did exist, but contained invalid
4277 data, a new version was generated, as expected. It was only the case of a
4278 non-existent file that was broken.
4280 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4281 with a change in libdomainkeys > 0.64.
4283 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4284 from DNS. If the selector record carries the flag, it now has
4285 precedence over the domain-wide flag.
4287 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4289 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4290 the use of a port name that isn't defined in /etc/services) occurred, the
4291 message was deferred as in a normal delivery, and thus remained on the
4292 spool, instead of being failed because of the mua_wrapper setting. This
4293 is now fixed, and I tidied up some of the mua_wrapper messages at the
4296 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4297 the messages in a hash of arrays rather than using individual hashes.
4298 This is a bit cleaner and results in dramatic memory savings, albeit
4299 at a slight CPU cost.
4301 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4302 as requested by Marc Sherman.
4304 SC/10 Eximstats: added histograms for user specified patterns as requested
4307 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4309 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4310 fopen() in the content-scanning modules that did not already have it.
4316 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4317 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4319 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4321 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4323 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4324 to test Sieve filters that use "vacation".
4326 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4327 that changes the way the GnuTLS parameters are stored in the cache file.
4328 The new format can be generated externally. For backward compatibility,
4329 if the data in the cache doesn't make sense, Exim assumes it has read an
4330 old-format file, and it generates new data and writes a new file. This
4331 means that you can't go back to an older release without removing the
4334 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4335 work if there are any delivery delays because "one_time" forces the
4336 parent to be marked "delivered", so its unseen clone is never tried
4337 again. For this reason, Exim now forbids the simultaneous setting of
4340 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4341 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4342 entry if you want to know the details. The fix, however, neglected to
4343 consider the case when local delivery batching is involved. The test for
4344 "previously delivered" was not happening when checking to see if an
4345 address could be batched with a previous (undelivered) one; under
4346 certain circumstances this could lead to multiple deliveries to the same
4349 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4350 in its include files, and this causes problems building Exim.
4352 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4353 header_syntax) but Exim was just ignoring anything given after a slash.
4354 In particular, this caused confusion with an attempt to use "verify =
4355 reverse_host_lookup/defer_ok". An error is now given when options are
4356 supplied for verify items that do not have them. (Maybe reverse_host_
4357 lookup should have a defer_ok option, but that's a different point.)
4359 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4360 defined by RFC 821) to 2048, because there were problems with some AUTH
4361 commands, and RFC 1869 says the size should be increased for extended
4362 SMTP commands that take arguments.
4364 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4367 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4368 "unknown" error; now it says that the functionality isn't in the binary.
4370 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4371 an address' error message when a string expansion fails (syntax or
4372 whatever). Otherwise the password may appear in the log. Following change
4373 PH/42 below, there is no longer a chance of it appearing in a bounce
4376 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4378 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4379 of its addresses was ever tried. (Bugzilla bug #2.)
4381 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4382 the result incorrectly in the debug output. (It correctly added a newline
4383 to what was transported.)
4385 TF/01 Added $received_time.
4387 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4388 commented out examples of how to interface to a virus scanner and to
4389 SpamAssassin. Also added commented examples of av_scanner and
4390 spamd_address settings.
4392 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4393 and controls are allowed in which ACLs. There were a couple of minor
4394 errors. Some of the entries in the conditions table (which is a table of
4395 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4396 negation of where the condition IS allowed.
4398 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4400 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4401 header file does not have a version number, so I've had to invent a new
4402 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4403 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4404 radiusclient), but it was contributed by a Radius user.
4406 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4407 files or directories, for OpenSSL.
4409 PH/19 When an Exim process that is running as root has to create an Exim log
4410 file, it does so in a subprocess that runs as exim:exim so as to get the
4411 ownership right at creation (otherwise, other Exim processes might see
4412 the file with the wrong ownership). There was no test for failure of this
4413 fork() call, which would lead to the process getting stuck as it waited
4414 for a non-existent subprocess. Forks do occasionally fail when resources
4415 run out. I reviewed all the other calls to fork(); they all seem to check
4418 PH/20 When checking for unexpected SMTP input at connect time (before writing
4419 the banner), Exim was not dealing correctly with a non-positive return
4420 from the read() function. If the client had disconnected by this time,
4421 the result was a log entry for a synchronization error with an empty
4422 string after "input=" when read() returned zero. If read() returned -1
4423 (an event I could not check), uninitialized data bytes were printed.
4424 There were reports of junk text (parts of files, etc) appearing after
4427 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4429 PH/22 Added support for macro redefinition, and (re)definition in between
4430 driver and ACL definitions.
4432 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4433 forgetting to use the resulting value; it was using the unexpanded value.
4435 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4436 hadn't been configured. The fix is from Juergen Kreileder, who
4437 understands it better than I do:
4439 "Here's what I see happening with three configured cyrus_sasl
4440 authenticators configured (plain, login, cram-md5):
4442 On startup auth_cyrus_sasl_init() gets called for each of these.
4443 This means three calls to sasl_listmech() without a specified mech_list.
4444 => SASL tests which mechs of all available mechs actually work
4445 => three warnings about OTP not working
4446 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4448 With the patch, sasl_listmech() also gets called three times. But now
4449 SASL's mech_list option is set to the server_mech specified in the the
4450 authenticator. Or in other words, the answer from sasl_listmech()
4451 gets limited to just the mech you're testing for (which is different
4453 => the return list contains just 'plain' or 'login', 'cram-md5' or
4454 nothing depending on the value of ob->server_mech.
4456 I've just tested the patch: Authentication still works fine,
4457 unavailable mechs specified in the exim configuration are still
4458 caught, and the auth.log warnings about OTP are gone."
4460 PH/25 When debugging is enabled, the contents of the command line are added
4461 to the debugging output, even when log_selector=+arguments is not
4464 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4465 answer is "GNU", and only if the return is "GNU/something" is the answer
4468 PH/27 $acl_verify_message is now set immediately after the failure of a
4469 verification in an ACL, and so is available in subsequent modifiers. In
4470 particular, the message can be preserved by coding like this:
4472 warn !verify = sender
4473 set acl_m0 = $acl_verify_message
4475 Previously, $acl_verify_message was set only while expanding "message"
4476 and "log_message" when a very denied access.
4478 PH/28 Modified OS/os.c-Linux with
4480 -#ifndef OS_LOAD_AVERAGE
4481 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4483 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4484 nomenclature these days.)
4486 PH/29 Installed patch from the Sieve maintainer that adds the options
4487 sieve_useraddress and sieve_subaddress to the redirect router.
4489 PH/30 In these circumstances:
4490 . Two addresses routed to the same list of hosts;
4491 . First host does not offer TLS;
4492 . First host accepts first address;
4493 . First host gives temporary error to second address;
4494 . Second host offers TLS and a TLS session is established;
4495 . Second host accepts second address.
4496 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4497 and peerdn, if requested) that were in fact used only for the second
4500 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4501 attention to any local part prefix or suffix that was matched by the
4502 router that accepted the address. It now behaves in the same way as it
4503 does for delivery: the affixes are removed from the local part unless
4504 rcpt_include_affixes is set on the transport.
4506 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4507 timeout during the DATA phase of an incoming message.
4509 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4510 applied a patch sanctioned by the Sieve maintainer.
4512 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4513 the uid or gid is negative. A case of a negative gid caused this to be
4514 noticed. The fix allows for either to be negative.
4516 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4517 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4520 PH/36 The change PH/12 above was broken. Fixed it.
4522 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4523 the grounds that routing the same address twice would always produce the
4524 same answer. This might have been true once, but it is certainly no
4525 longer true now. Routing a child address may depend on the previous
4526 routing that produced that child. Some complicated redirection strategies
4527 went wrong when messages had multiple recipients, and made Exim's
4528 behaviour dependent on the order in which the addresses were given.
4530 I have moved the duplicate checking until after the routing is complete.
4531 Exim scans the addresses that are assigned to local and remote
4532 transports, and removes any duplicates. This means that more work will be
4533 done, as duplicates will always all be routed, but duplicates are
4534 presumably rare, so I don't expect this is of any significance.
4536 For deliveries to pipes, files, and autoreplies, the duplicate checking
4537 still happens during the routing process, since they are not going to be
4540 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4541 It corrects a timeout issue with spamd. This is Ian's comment: "The
4542 background is that sometimes spamd either never reads data from a
4543 connection it has accepted, or it never writes response data. The exiscan
4544 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4545 blindly assumes that writes won't block so it may never time out."
4547 PH/39 Allow G after quota size as well as K and M.
4549 PH/40 The value set for $authenticated_id in an authenticator may not contain
4550 binary zeroes or newlines because the value is written to log lines and
4551 to spool files. There was no check on this. Now the value is run through
4552 the string_printing() function so that such characters are converted to
4553 printable escape sequences.
4555 PH/41 $message_linecount is a new variable that contains the total number of
4556 lines in the message. Compare $body_linecount, which is the count for the
4559 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4560 bounce and delay warning messages, except in certain special cases, which
4563 (a) An SMTP error message from a remote host;
4564 (b) A message specified in a :fail: redirection;
4565 (c) A message specified in a "fail" command in a system filter;
4566 (d) A message specified in a FAIL return from the queryprogram router;
4567 (e) A message specified by the cannot_route_message router option.
4569 In these cases only, Exim does include the error details in bounce and
4570 warning messages. There are also a few cases where bland messages such
4571 as "unrouteable address" or "local delivery error" are given.
4573 PH/43 $value is now also set for the "else" part of a ${run expansion.
4575 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4576 being worked on, but at least Exim now implements the latest version to
4579 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4580 process to complete was treated as a delivery failure, a timeout while
4581 writing the message to the pipe was logged, but erroneously treated as a
4582 successful delivery. Such timeouts include transport filter timeouts. For
4583 consistency with the overall process timeout, these timeouts are now
4584 treated as errors, giving rise to delivery failures by default. However,
4585 there is now a new Boolean option for the pipe transport called
4586 timeout_defer, which, if set TRUE, converts the failures into defers for
4587 both kinds of timeout. A transport filter timeout is now identified in
4590 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4591 systems where "make" and "gmake" are different, calling "gmake" at top
4592 level broke things. I've arranged for the value of $(MAKE) to be passed
4593 from the Makefile to this script so that it can call the same version of
4597 A note about Exim versions 4.44 and 4.50
4598 ----------------------------------------
4600 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4601 changes of various kinds. As a consequence, a big documentation update was
4602 needed. This delayed the release for rather longer than seemed good, especially
4603 in the light of a couple of (minor) security issues. Therefore, the changes
4604 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4605 release. So 4.44 and 4.50 are in effect two different branches that both start
4608 I have left the 4.50 change log unchanged; it contains all the changes since
4609 4.43. The change log for 4.44 is below; many of its items are identical to
4610 those for 4.50. This seems to be the most sensible way to preserve the
4611 historical information.
4617 1. Minor wording change to the doc/README.SIEVE file.
4619 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4620 computation of the current number of files was incorrect.
4622 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4623 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4626 4. Give more explanation in the error message when the command for a transport
4627 filter fails to execute.
4629 5. There are several places where Exim runs a non-Exim command in a
4630 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4631 was being done only for the command run by the queryprogram router. It is
4632 now done for all such subprocesses. The other cases are: ${run, transport
4633 filters, and the commands run by the lmtp and pipe transports.
4635 6. Added CONFIGURE_GROUP build-time option.
4637 7. Some older OS have a limit of 256 on the maximum number of file
4638 descriptors. Exim was using setrlimit() to set 1000 as a large value
4639 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4640 systems. I've change it so that if it can't get 1000, it tries for 256.
4642 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4643 was an oversight, and furthermore, ever since the addition of extra
4644 controls (e.g. 4.43/32), the checks on when to allow different forms of
4645 "control" were broken. There should now be diagnostics for all cases when a
4646 control that does not make sense is encountered.
4648 9. Added the /retain_sender option to "control=submission".
4650 10. $recipients is now available in the predata ACL (oversight).
4652 11. Tidy the search cache before the fork to do a delivery from a message
4653 received from the command line. Otherwise the child will trigger a lookup
4654 failure and thereby defer the delivery if it tries to use (for example) a
4655 cached ldap connection that the parent has called unbind on.
4657 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4658 of $address_data from the recipient verification was clobbered by the
4659 sender verification.
4661 13. The value of address_data from a sender verification is now available in
4662 $sender_address_data in subsequent conditions in the ACL statement.
4664 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4666 15. Added a new option "connect=<time>" to callout options, to set a different
4669 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4670 was its contents. (It was OK if the option was not defined at all.)
4672 17. A "Completed" log line is now written for messages that are removed from
4673 the spool by the -Mrm option.
4675 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4676 information about exactly what failed.
4678 19. Added -dd to debug only the daemon process.
4680 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4681 handles timeouts, both on the server side and network timeouts. Renamed the
4682 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4684 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4685 It is now set to "smtps".
4687 22. $host_address is now set to the target address during the checking of
4688 ignore_target_hosts.
4690 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4691 being passed; this would have caused $sender_host_name to have been used if
4692 matching the list had actually called for a host name (not very likely,
4693 since this list is usually IP addresses). A host name is now passed as
4696 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4697 code that specifies a non-restarting handler (typically sigaction() in
4698 modern systems) in an attempt to fix a rare and obscure crash bug.
4700 25. Narrowed the window for a race in the daemon that could cause it to ignore
4701 SIGCHLD signals. This is not a major problem, because they are used only to
4702 wake it up if nothing else does.
4704 26. A malformed maildirsize file could cause Exim to calculate negative values
4705 for the mailbox size or file count. Odd effects could occur as a result.
4706 The maildirsize information is now recalculated if the size or filecount
4709 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4710 support for a long time. Removed HAVE_SYS_VFS_H.
4712 28. Installed the latest version of exipick from John Jetmore.
4714 29. In an address list, if the pattern was not a regular expression, an empty
4715 subject address (from a bounce message) matched only if the pattern was an
4716 empty string. Non-empty patterns were not even tested. This was the wrong
4717 because it is perfectly reasonable to use an empty address as part of a
4718 database query. An empty address is now tested by patterns that are
4719 lookups. However, all the other forms of pattern expect the subject to
4720 contain a local part and a domain, and therefore, for them, an empty
4721 address still always fails if the pattern is not itself empty.
4723 30. Exim went into a mad DNS loop when attempting to do a callout where the
4724 host was specified on an smtp transport, and looking it up yielded more
4725 than one IP address.
4727 31. Re-factored the code for checking spool and log partition space into a
4728 function that finds that data and another that does the check. The former
4729 is then used to implement four new variables: $spool_space, $log_space,
4730 $spool_inodes, and $log_inodes.
4732 32. The RFC2047 encoding function was originally intended for short strings
4733 such as real names; it was not keeping to the 75-character limit for
4734 encoded words that the RFC imposes. It now respects the limit, and
4735 generates multiple encoded words if necessary. To be on the safe side, I
4736 have increased the buffer size for the ${rfc2047: expansion operator from
4739 33. It is now permitted to omit both strings after an "if" condition; if the
4740 condition is true, the result is "true". As before, when the second string
4741 is omitted, a false condition yields an empty string. This makes it less
4742 cumbersome to write custom ACL and router conditions.
4744 34. Failure to deliver a bounce message always caused it to be frozen, even if
4745 there was an errors_to setting on the router. The errors_to setting is now
4748 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4749 canonical form (fully expanded) before being placed in
4750 $sender_host_address.
4752 36. The table in the code that translates DNS record types into text (T_A to
4753 "A" for instance) was missing entries for NS and CNAME. It is just possible
4754 that this could have caused confusion if both these types were looked up
4755 for the same domain, because the text type is used as part of Exim's
4756 per-process caching. But the chance of anyone hitting this buglet seems
4759 37. The dnsdb lookup has been extended in a number of ways.
4761 (1) There is a new type, "zns", which walks up the domain tree until it
4762 finds some nameserver records. It should be used with care.
4764 (2) There is a new type, "mxh", which is like "mx" except that it returns
4765 just the host names, not the priorities.
4767 (3) It is now possible to give a list of domains (or IP addresses) to be
4768 looked up. The behaviour when one of the lookups defers can be
4769 controlled by a keyword.
4771 (4) It is now possible to specify the separator character for use when
4772 multiple records are returned.
4774 38. The dnslists ACL condition has been extended: it is now possible to supply
4775 a list of IP addresses and/or domains to be looked up in a particular DNS
4778 39. Added log_selector=+queue_time_overall.
4780 40. When running the queue in the test harness, wait just a tad after forking a
4781 delivery process, to get repeatability of debugging output.
4783 41. Include certificate and key file names in error message when GnuTLS fails
4784 to set them up, because the GnuTLS error message doesn't include the name
4785 of the failing file when there is a problem reading it.
4787 42. Allow both -bf and -bF in the same test run.
4789 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4791 44. The "Exiscan patch" is now merged into the mainline Exim source.
4793 45. Sometimes the final signoff response after QUIT could fail to get
4794 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4795 < 0 before doing a fflush(). This bug looks as though it goes back to the
4796 introduction of TLS in release 3.20, but "sometimes" must have been rare
4797 because the tests only now provoked it.
4799 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4800 (this can affect the format of dates).
4802 47. exim_tidydb, when checking for the continued existence of a message for
4803 which it has found a message-specific retry record, was not finding
4804 messages that were in split spool directories. Consequently, it was
4805 deleting retry records that should have stayed in existence.
4807 48. Steve fixed some bugs in eximstats.
4809 49. The SPA authentication driver was not abandoning authentication and moving
4810 on to the next authenticator when an expansion was forced to fail,
4811 contradicting the general specification for all authenticators. Instead it
4812 was generating a temporary error. It now behaves as specified.
4814 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4815 (the order specifies the preference for clients). The order is now AES256,
4816 AES128, 3DES, ARCFOUR128.
4818 51. Small patch to Sieve code - explicitly set From: when generating an
4821 52. Exim crashed if a remote delivery caused a very long error message to be
4822 recorded - for instance if somebody sent an entire SpamAssassin report back
4823 as a large number of 550 error lines. This bug was coincidentally fixed by
4824 increasing the size of one of Exim's internal buffers (big_buffer) that
4825 happened as part of the Exiscan merge. However, to be on the safe side, I
4826 have made the code more robust (and fixed the comments that describe what
4829 53. Now that there can be additional text after "Completed" in log lines (if
4830 the queue_time_overall log selector is set), a one-byte patch to exigrep
4831 was needed to allow it to recognize "Completed" as not the last thing in
4834 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4835 patch that reportedly fixes this has been added. I am not expert enough to
4836 create a test for it. This is what the patch creator wrote:
4838 "I found a little strange behaviour of ldap code when working with
4839 Windows 2003 AD Domain, where users was placed in more than one
4840 Organization Units. When I tried to give exim partial DN, the exit code
4841 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4842 But simultaneously result of request was absolutely normal ldap result,
4843 so I produce this patch..."
4845 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4846 REFERENCE, so I have modified the code to exclude the patch when that macro
4849 55. Some experimental protocols are using DNS PTR records for new purposes. The
4850 keys for these records are domain names, not reversed IP addresses. The
4851 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4852 leaves it alone. Component reversal etc. now happens only for IP addresses.
4855 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4857 57. Double the size of the debug message buffer (to 2048) so that more of very
4858 long debug lines gets shown.
4860 58. The exicyclog utility now does better if the number of log files to keep
4861 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4863 59. Two changes related to the smtp_active_hostname option:
4865 (1) $smtp_active_hostname is now available as a variable.
4866 (2) The default for smtp_banner uses $smtp_active_hostname instead
4867 of $primary_hostname.
4869 60. The host_aton() function is supposed to be passed a string that is known
4870 to be a valid IP address. However, in the case of IPv6 addresses, it was
4871 not checking this. This is a hostage to fortune. Exim now panics and dies
4872 if the condition is not met. A case was found where this could be provoked
4873 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4874 components; fortuitously, this particular loophole had already been fixed
4875 by change 4.50/55 above.
4877 If there are any other similar loopholes, the new check in host_aton()
4878 itself should stop them being exploited. The report I received stated that
4879 data on the command line could provoke the exploit when Exim was running as
4880 exim, but did not say which command line option was involved. All I could
4881 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4882 running as the user.
4885 61. There was a buffer overflow vulnerability in the SPA authentication code
4886 (which came originally from the Samba project). I have added a test to the
4887 spa_base64_to_bits() function which I hope fixes it.
4890 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4891 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4893 63. The daemon start-up calls getloadavg() while still root for those OS that
4894 need the first call to be done as root, but it missed one case: when
4895 deliver_queue_load_max is set with deliver_drop_privilege. This is
4896 necessary for the benefit of the queue runner, because there is no re-exec
4897 when deliver_drop_privilege is set.
4899 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4900 This has been fixed.
4902 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4903 was in use, was not putting the data itself into the right store pool;
4904 consequently, it could be overwritten for a subsequent message in the same
4905 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4908 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4910 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4911 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4912 the function were treating the return as a boolean value, which happened to
4913 work because 0=false and not-0=true, but is not correct code.
4915 68. The host_aton() function was not handling scoped IPv6 addresses (those
4916 with, for example, "%eth0" on the end) correctly.
4918 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4919 negated items (that is, ~something) in unsigned ints. Some compilers
4920 apparently mutter when there is no cast.
4922 70. If an address verification called from an ACL failed, and did not produce a
4923 user-specific message (i.e. there was only a "system" message), nothing was
4924 put in $acl_verify_message. In this situation, it now puts the system
4927 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4928 session; change 4.31/43 added the unwanted input to the log line - except
4929 that it did not do this in the start of session case. It now does.
4931 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4932 This is wrong and can cause the other end to generate a synchronization
4933 error if it is another Exim or anything else that does the synchronization
4934 check. A QUIT command is no longer sent after a timeout.
4936 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4937 during host lookups.
4939 74. The defer_ok option of callout verification was not working if it was used
4940 when verifying addresses in header lines, that is, for this case:
4942 verify = header_sender/callout=defer_ok
4944 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4945 those file descriptors could be used for SMTP connections. If anything
4946 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4947 could be sent to the SMTP client, causing chaos. The daemon now opens
4948 stdin, stdout, and stderr to /dev/null when it puts itself into the
4951 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4952 log by default. The user can override this with suitable Perl magic.
4954 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4955 the log message when discard triggers, was not working for the DATA ACL or
4956 for the non-SMTP ACL.
4958 78. Error message wording change in sieve.c.
4960 79. If smtp_accept_max_per_host was set, the number of connections could be
4961 restricted to fewer than expected, because the daemon was trying to set up
4962 a new connection before checking whether the processes handling previous
4963 connections had finished. The check for completed processes is now done
4964 earlier. On busy systems, this bug wouldn't be noticed because something
4965 else would have woken the daemon, and it would have reaped the completed
4968 80. If a message was submitted locally by a user whose login name contained one
4969 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4970 It caused a spool format error. I have fixed the spool reading code. A
4971 related problem was that the "from" clause in the Received: line became
4972 illegal because of the space(s). It is now covered by ${quote_local_part.
4974 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4977 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4979 83. Patch from Sieve maintainer to fix unterminated string problem in
4980 "vacation" handling.
4982 84. Some minor changes to the Linux configuration files to help with other
4983 OS variants using glibc.
4985 85. One more patch for Sieve to update vacation handling to latest spec.
4988 ----------------------------------------------------
4989 See the note above about the 4.44 and 4.50 releases.
4990 ----------------------------------------------------
4996 1. Change 4.43/35 introduced a bug that caused file counts to be
4997 incorrectly computed when quota_filecount was set in an appendfile
5000 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5001 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5004 3. Give more explanation in the error message when the command for a transport
5005 filter fails to execute.
5007 4. There are several places where Exim runs a non-Exim command in a
5008 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5009 was being done only for the command run by the queryprogram router. It is
5010 now done for all such subprocesses. The other cases are: ${run, transport
5011 filters, and the commands run by the lmtp and pipe transports.
5013 5. Some older OS have a limit of 256 on the maximum number of file
5014 descriptors. Exim was using setrlimit() to set 1000 as a large value
5015 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5016 systems. I've change it so that if it can't get 1000, it tries for 256.
5018 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5019 was an oversight, and furthermore, ever since the addition of extra
5020 controls (e.g. 4.43/32), the checks on when to allow different forms of
5021 "control" were broken. There should now be diagnostics for all cases when a
5022 control that does not make sense is encountered.
5024 7. $recipients is now available in the predata ACL (oversight).
5026 8. Tidy the search cache before the fork to do a delivery from a message
5027 received from the command line. Otherwise the child will trigger a lookup
5028 failure and thereby defer the delivery if it tries to use (for example) a
5029 cached ldap connection that the parent has called unbind on.
5031 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5032 of $address_data from the recipient verification was clobbered by the
5033 sender verification.
5035 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5036 was its contents. (It was OK if the option was not defined at all.)
5038 11. A "Completed" log line is now written for messages that are removed from
5039 the spool by the -Mrm option.
5041 12. $host_address is now set to the target address during the checking of
5042 ignore_target_hosts.
5044 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5045 being passed; this would have caused $sender_host_name to have been used if
5046 matching the list had actually called for a host name (not very likely,
5047 since this list is usually IP addresses). A host name is now passed as
5050 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5051 code that specifies a non-restarting handler (typically sigaction() in
5052 modern systems) in an attempt to fix a rare and obscure crash bug.
5054 15. Narrowed the window for a race in the daemon that could cause it to ignore
5055 SIGCHLD signals. This is not a major problem, because they are used only to
5056 wake it up if nothing else does.
5058 16. A malformed maildirsize file could cause Exim to calculate negative values
5059 for the mailbox size or file count. Odd effects could occur as a result.
5060 The maildirsize information is now recalculated if the size or filecount
5063 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5064 support for a long time. Removed HAVE_SYS_VFS_H.
5066 18. Updated exipick to current release from John Jetmore.
5068 19. Allow an empty sender to be matched against a lookup in an address list.
5069 Previously the only cases considered were a regular expression, or an
5072 20. Exim went into a mad DNS lookup loop when doing a callout where the
5073 host was specified on the transport, if the DNS lookup yielded more than
5076 21. The RFC2047 encoding function was originally intended for short strings
5077 such as real names; it was not keeping to the 75-character limit for
5078 encoded words that the RFC imposes. It now respects the limit, and
5079 generates multiple encoded words if necessary. To be on the safe side, I
5080 have increased the buffer size for the ${rfc2047: expansion operator from
5083 22. Failure to deliver a bounce message always caused it to be frozen, even if
5084 there was an errors_to setting on the router. The errors_to setting is now
5087 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5088 canonical form (fully expanded) before being placed in
5089 $sender_host_address.
5091 24. Updated eximstats to version 1.33
5093 25. Include certificate and key file names in error message when GnuTLS fails
5094 to set them up, because the GnuTLS error message doesn't include the name
5095 of the failing file when there is a problem reading it.
5097 26. Expand error message when OpenSSL has problems setting up cert/key files.
5100 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5101 (this can affect the format of dates).
5103 28. exim_tidydb, when checking for the continued existence of a message for
5104 which it has found a message-specific retry record, was not finding
5105 messages that were in split spool directories. Consequently, it was
5106 deleting retry records that should have stayed in existence.
5108 29. eximstats updated to version 1.35
5109 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5110 1.35 - bugfix such that pie charts by volume are generated correctly
5112 30. The SPA authentication driver was not abandoning authentication and moving
5113 on to the next authenticator when an expansion was forced to fail,
5114 contradicting the general specification for all authenticators. Instead it
5115 was generating a temporary error. It now behaves as specified.
5117 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5118 (the order specifies the preference for clients). The order is now AES256,
5119 AES128, 3DES, ARCFOUR128.
5121 31. Small patch to Sieve code - explicitly set From: when generating an
5124 32. Exim crashed if a remote delivery caused a very long error message to be
5125 recorded - for instance if somebody sent an entire SpamAssassin report back
5126 as a large number of 550 error lines. This bug was coincidentally fixed by
5127 increasing the size of one of Exim's internal buffers (big_buffer) that
5128 happened as part of the Exiscan merge. However, to be on the safe side, I
5129 have made the code more robust (and fixed the comments that describe what
5132 33. Some experimental protocols are using DNS PTR records for new purposes. The
5133 keys for these records are domain names, not reversed IP addresses. The
5134 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5135 leaves it alone. Component reversal etc. now happens only for IP addresses.
5138 34. The host_aton() function is supposed to be passed a string that is known
5139 to be a valid IP address. However, in the case of IPv6 addresses, it was
5140 not checking this. This is a hostage to fortune. Exim now panics and dies
5141 if the condition is not met. A case was found where this could be provoked
5142 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5143 components; fortuitously, this particular loophole had already been fixed
5144 by change 4.50/55 or 4.44/33 above.
5146 If there are any other similar loopholes, the new check in host_aton()
5147 itself should stop them being exploited. The report I received stated that
5148 data on the command line could provoke the exploit when Exim was running as
5149 exim, but did not say which command line option was involved. All I could
5150 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5151 running as the user.
5154 35. There was a buffer overflow vulnerability in the SPA authentication code
5155 (which came originally from the Samba project). I have added a test to the
5156 spa_base64_to_bits() function which I hope fixes it.
5159 36. The daemon start-up calls getloadavg() while still root for those OS that
5160 need the first call to be done as root, but it missed one case: when
5161 deliver_queue_load_max is set with deliver_drop_privilege. This is
5162 necessary for the benefit of the queue runner, because there is no re-exec
5163 when deliver_drop_privilege is set.
5165 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5166 was in use, was not putting the data itself into the right store pool;
5167 consequently, it could be overwritten for a subsequent message in the same
5168 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5171 38. Sometimes the final signoff response after QUIT could fail to get
5172 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5173 < 0 before doing a fflush(). This bug looks as though it goes back to the
5174 introduction of TLS in release 3.20, but "sometimes" must have been rare
5175 because the tests only now provoked it.
5181 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5182 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5183 Now it returns an integer. A number of calls were still expecting a T/F
5184 return. Fortuitously, in all cases, the tests worked in OK situations,
5185 which is the norm. However, things would have gone wrong on any write
5186 failures on the smtp file descriptor. This function is used when sending
5187 messages over SMTP and also when doing verify callouts.
5189 2. When Exim is called to do synchronous delivery of a locally submitted
5190 message (the -odf or -odi options), it no longer closes stderr before doing
5193 3. Implemented the mua_wrapper option.
5195 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5197 5. Implemented the functions header_remove(), header_testname(),
5198 header_add_at_position(), and receive_remove_recipient(), and exported them
5201 6. If an ACL "warn" statement specified the addition of headers, Exim already
5202 inserted X-ACL-Warn: at the start if there was no header name. However, it
5203 was not making this test for the second and subsequent header lines if
5204 there were newlines in the string. This meant that an invalid header could
5205 be inserted if Exim was badly configured.
5207 7. Allow an ACL "warn" statement to add header lines at the start or after all
5208 the Received: headers, as well as at the end.
5210 8. Added the rcpt_4xx retry error code.
5212 9. Added postmaster_mailfrom=xxx to callout verification option.
5214 10. Added mailfrom=xxxx to the callout verification option, for verify=
5217 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5218 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5220 12. Inserted some casts to stop certain compilers warning when using pointer
5221 differences as field lengths or precisions in printf-type calls (mostly
5222 affecting debugging statements).
5224 13. Added optional readline() support for -be (dynamically loaded).
5226 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5227 same clock tick as a message's arrival, so that its received time was the
5228 same as the "first fail" time on the retry record, and that message
5229 remained on the queue past the ultimate address timeout, every queue runner
5230 would try a delivery (because it was past the ultimate address timeout) but
5231 after another failure, the ultimate address timeout, which should have then
5232 bounced the address, did not kick in. This was a "< instead of <=" error;
5233 in most cases the first failure would have been in the next clock tick
5234 after the received time, and all would be well.
5236 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5237 being recognized when the domain list was tested by the match_domain
5238 condition in an expansion string.
5240 16. Added the ${str2b64: operator.
5242 17. Exim was always calling setrlimit() to set a large limit for the number of
5243 processes, without checking whether the existing limit was already
5244 adequate. (It did check for the limit on file descriptors.) Furthermore,
5245 errors from getrlimit() and setrlimit() were being ignored. Now they are
5246 logged to the main and panic logs, but Exim does carry on, to try to do its
5247 job under whatever limits there are.
5249 18. Imported PCRE 5.0.
5251 19. Trivial typo in log message " temporarily refused connection" (the leading
5254 20. If the log selector return_path_on_delivery was set and an address was
5255 redirected to /dev/null, the delivery process crashed because it assumed
5256 that a return path would always be set for a "successful" delivery. In this
5257 case, the whole delivery is bypassed as an optimization, and therefore no
5260 21. Internal re-arrangement: the function for sending a challenge and reading
5261 a response while authentication was assuming a zero-terminated challenge
5262 string. It's now changed to take a pointer and a length, to allow for
5263 binary data in such strings.
5265 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5267 23. Exim was not respecting finduser_retries when seeking the login of the
5268 uid under which it was called; it was always trying 10 times. (The default
5269 setting of finduser_retries is zero.) Also, it was sleeping after the final
5270 failure, which is pointless.
5272 24. Implemented tls_on_connect_ports.
5274 25. Implemented acl_smtp_predata.
5276 26. If the domain in control=submission is set empty, Exim assumes that the
5277 authenticated id is a complete email address when it generates From: or
5278 Sender: header lines.
5280 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5281 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5282 chown and chgrp in /bin and hostname in /usr/bin.
5284 28. Exim was keeping the "process log" file open after each use, just as it
5285 does for the main log. This opens the possibility of it remaining open for
5286 long periods when the USR1 signal hits a daemon. Occasional processlog
5287 errors were reported, that could have been caused by this. Anyway, it seems
5288 much more sensible not to leave this file open at all, so that is what now
5291 29. The long-running daemon process does not normally write to the log once it
5292 has entered its main loop, and it closes the log before doing so. This is
5293 so that log files can straightforwardly be renamed and moved. However,
5294 there are a couple of unusual error situations where the daemon does write
5295 log entries, and I had neglected to close the log afterwards.
5297 30. The text of an SMTP error response that was received during a remote
5298 delivery was being truncated at 512 bytes. This is too short for some of
5299 the long messages that one sometimes sees. I've increased the limit to
5302 31. It is now possible to make retry rules that apply only when a message has a
5303 specific sender, in particular, an empty sender.
5305 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5306 it possible to be selective about when SMTP synchronization is enforced.
5308 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5310 32. Implemented hosts_connection_nolog.
5312 33. Added an ACL for QUIT.
5314 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5317 35. Added mailbox_size and mailbox_filecount to appendfile.
5319 36. Added control = no_multiline_responses to ACLs.
5321 37. There was a bug in the logic of the code that waits for the clock to tick
5322 in the case where the clock went backwards by a substantial amount such
5323 that the microsecond fraction of "now" was more than the microsecond
5324 fraction of "then" (but the whole seconds number was less).
5326 38. Added support for the libradius Radius client library this is found on
5327 FreeBSD (previously only the radiusclient library was supported).
5333 1. When certain lookups returned multiple values in the form name=value, the
5334 quoting of the values was not always being done properly. Specifically:
5335 (a) If the value started with a double quote, but contained no whitespace,
5337 (b) If the value contained whitespace other than a space character (i.e.
5338 tabs or newlines or carriage returns) it was not quoted.
5339 This fix has been applied to the mysql and pgsql lookups by writing a
5340 separate quoting function and calling it from the lookup code. The fix
5341 should probably also be applied to nisplus, ibase and oracle lookups, but
5342 since I cannot test any of those, I have not disturbed their existing code.
5344 2. A hit in the callout cache for a specific address caused a log line with no
5345 reason for rejecting RCPT. Now it says "Previous (cached) callout
5346 verification failure".
5348 3. There was an off-by-one bug in the queryprogram router. An over-long
5349 return line was truncated at 256 instead of 255 characters, thereby
5350 overflowing its buffer with the terminating zero. As well as fixing this, I
5351 have increased the buffer size to 1024 (and made a note to document this).
5353 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5354 when Exim is waiting for an SMTP response from a remote server, Exim
5355 restarts its select() call on the socket, thereby resetting its timeout.
5356 This is not a problem when such interrupts are rare. Somebody set up a cron
5357 job to run exiwhat every 2 minutes, which is less than the normal select()
5358 timeout (5 or 10 minutes). This meant that the select() timeout never
5359 kicked in because it was always reset. I have fixed this by comparing the
5360 time when an interrupt arrives with the time at the start of the first call
5361 to select(). If more time than the timeout has elapsed, the interrupt is
5362 treated as a timeout.
5364 5. Some internal re-factoring in preparation for the addition of Sieve
5365 extensions (by MH). In particular, the "personal" test is moved to a
5366 separate function, and given an option for scanning Cc: and Bcc: (which is
5367 not set for Exim filters).
5369 6. When Exim created an email address using the login of the caller as the
5370 local part (e.g. when creating a From: or Sender: header line), it was not
5371 quoting the local part when it contained special characters such as @.
5373 7. Installed new OpenBSD configuration files.
5375 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5376 try to make them clearer.
5378 9. Callout options, other than the timeout value, were being ignored when
5379 verifying sender addresses in header lines. For example, when using
5381 verify = header_sender/callout=no_cache
5383 the cache was (incorrectly) being used.
5385 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5386 only the Cygwin environment.
5388 11. When return_path_on_delivery was set as a log selector, if different remote
5389 addresses in the same message used different return paths and parallel
5390 remote delivery occurred, the wrong values would sometimes be logged.
5391 (Whenever a remote delivery process finished, the return path value from
5392 the most recently started remote delivery process was logged.)
5394 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5395 lines when AUTH and/or TLS are in use. This is the "received protocol"
5396 field. Exim used to use "asmtp" for authenticated SMTP, without any
5397 indication (in the protocol name) for TLS use. Now it follows the RFC and
5398 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5399 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5400 names appear in log lines as well as in Received: header lines.
5402 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5403 extensions, and comparison tests, and to fix some bugs.
5405 14. Changes to the "personal" filter test:
5407 (1) The test was buggy in that it was just doing the equivalent of
5408 "contains" tests on header lines. For example, if a user's address was
5409 anne@some.where, the "personal" test would incorrectly be true for
5411 To: susanne@some.where
5413 This test is now done by extracting each address from the header in turn,
5414 and checking the entire address. Other tests that are part of "personal"
5415 are now done using regular expressions (for example, to check local parts
5416 of addresses in From: header lines).
5418 (2) The list of non-personal local parts in From: addresses has been
5419 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5420 taken from the Sieve specification recommendations.
5422 (3) If the message contains any header line starting with "List-" it is
5423 treated as non-personal.
5425 (4) The test for "circular" in the Subject: header line has been removed
5426 because it now seems ill-conceived.
5428 15. Minor typos in src/EDITME comments corrected.
5430 16. Installed latest exipick from John Jetmore.
5432 17. If headers_add on a router specified a text string that was too long for
5433 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5434 of string_sprintf() is now avoided.
5436 18. $message_body_size was not set (it was always zero) when running the DATA
5437 ACL and the local_scan() function.
5439 19. For the "mail" command in an Exim filter, no default was being set for
5440 the once_repeat time, causing a random time value to be used if "once" was
5441 specified. (If the value happened to be <= 0, no repeat happened.) The
5442 default is now 0s, meaning "never repeat". The "vacation" command was OK
5443 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5444 (I found it when inspecting the code).
5446 20. There is now an overall timeout for performing a callout verification. It
5447 defaults to 4 times the callout timeout, which applies to individual SMTP
5448 commands during the callout. The overall timeout applies when there is more
5449 than one host that can be tried. The timeout is checked before trying the
5450 next host. This prevents very long delays if there are a large number of
5451 hosts and all are timing out (e.g. when the network connections are timing
5452 out). The value of the overall timeout can be changed by specifying an
5453 additional sub-option for "callout", called "maxwait". For example:
5455 verify = sender/callout=5s,maxwait=20s
5457 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5458 to the end before writing, but this should make it even safer).
5460 22. Exim was forgetting that it had advertised PIPELINING for the second and
5461 subsequent messages on an SMTP connection. It was also not resetting its
5462 memory on STARTTLS and an internal HELO.
5464 23. When Exim logs an SMTP synchronization error within a session, it now
5465 records whether PIPELINING has been advertised or not.
5467 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5468 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5469 rather than long int.
5471 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5473 26. Added the never_mail option to autoreply.
5479 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5480 crash if the getsockname() call failed; this can happen if a connection is
5481 closed very soon after it is established. The problem was simply in the
5482 order in which certain operations were done, causing Exim to try to write
5483 to the SMTP stream before it had set up the file descriptor. The bug has
5484 been fixed by making things happen in the correct order.
5490 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5491 before the connection was closed, thus losing the rejection response.
5493 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5494 some early Solaris releases, but causes trouble in current releases where
5495 socklen_t is defined.
5497 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5500 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5503 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5504 that releasing the top of it at the end releases what was used for sub-
5505 expansions (unless the block got too big). However, discard this block if
5506 the first thing is a variable or header, so that we can use its block when
5507 it is dynamic (useful for very large $message_headers, for example).
5509 6. Lookups now cache *every* query, not just the most recent. A new, separate
5510 store pool is used for this. It can be recovered when all lookup caches are
5511 flushed. Lookups now release memory at the end of their result strings.
5512 This has involved some general refactoring of the lookup sources.
5514 7. Some code has been added to the store_xxx() functions to reduce the amount
5515 of flapping under certain conditions.
5517 8. log_incoming_interface used to affect only the <= reception log lines. Now
5518 it causes the local interface and port to be added to several more SMTP log
5519 lines, for example "SMTP connection from", and rejection lines.
5521 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5523 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5525 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5526 could be overwritten at the end of the current message (or the start of a
5527 new message if it was set in a HELO ACL). The value is now preserved for
5528 the duration of the SMTP connection.
5530 12. If a transport had a headers_rewrite setting, and a matching header line
5531 contained an unqualified address, that address was qualified, even if it
5532 did not match any rewriting rules. The underlying bug was that the values
5533 of the flags that permit the existence of unqualified sender and recipient
5534 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5535 non-local messages, and by -bnq for local messages) were not being
5536 preserved with the message after it was received.
5538 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5539 "next input=" as part of the text comprising the host identity instead of
5540 the correct text. The code was using the same buffer for two different
5541 strings. However, depending on which order the printing function evaluated
5542 its arguments, the bug did not always show up. Under Linux, for example, my
5543 test suite worked just fine.
5545 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5546 doesn't work with some older versions of Perl. It has been changed to "my",
5547 which in any case is probably the better facility to use.
5549 15. A really picky compiler found some instances of statements for creating
5550 error messages that either had too many or two few arguments for the format
5553 16. The size of the buffer for calls to the DNS resolver has been increased
5554 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5555 for addresses that have a lot of PTR records. This alleviates a problem; it
5556 does not fully solve it.
5558 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5559 buffer now truncates the list and logs the incident, which is the same
5560 action as happens when Exim is looking up a host name and its aliases.
5561 Previously in this situation something unpredictable would happen;
5562 sometimes it was "internal error: store_reset failed".
5564 18. If a server dropped the connection unexpectedly when an Exim client was
5565 using GnuTLS and trying to read a response, the client delivery process
5566 crashed while trying to generate an error log message.
5568 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5569 string, for example:
5571 warn message = H1: something\nH2: something
5573 the text was added as a single header line from Exim's point of view
5574 though it ended up OK in the delivered message. However, searching for the
5575 second and subsequent header lines using $h_h2: did not work. This has been
5576 fixed. Similarly, if a system filter added multiple headers in this way,
5577 the routers could not see them.
5579 20. Expanded the error message when iplsearch is called with an invalid key to
5580 suggest using net-iplsearch in a host list.
5582 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5583 ACLs are no longer actually imposed (and a message to that effect is
5586 22. If a "gecos" field in a passwd entry contained escaped characters, in
5587 particular, if it contained a \" sequence, Exim got it wrong when building
5588 a From: or a Sender: header from that name. A second bug also caused
5589 incorrect handling when an unquoted " was present following a character
5590 that needed quoting.
5592 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5593 was not being matched caselessly.
5595 24. Arranged for all hyphens in the exim.8 source to be escaped with
5598 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5599 properly. Recipient callout cache records were still being keyed to include
5600 the sender, even when use_sender was set false. This led to far more
5601 callouts that were necessary. The sender is no longer included in the key
5602 when use_sender is false.
5604 26. Added "control = submission" modifier to ACLs.
5606 27. Added the ${base62d: operator to decode base 62 numbers.
5608 28. dnsdb lookups can now access SRV records.
5610 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5611 the configuration file.
5613 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5614 (-v) mode. This makes the output for a verbose queue run more intelligible.
5616 31. Added a use_postmaster feature to recipient callouts.
5618 32. Added the $body_zerocount variable, containing the number of binary zero
5619 bytes in the message body.
5621 33. The time of last modification of the "new" subdirectory is now used as the
5622 "mailbox time last read" when there is a quota error for a maildir
5625 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5627 35. Added +ignore_unknown as a special item in host lists.
5629 36. Code for decoding IPv6 addresses in host lists is now included, even if
5630 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5631 address was recognized as an IP address, but was then not correctly decoded
5632 into binary, causing unexpected and incorrect effects when compared with
5639 1. Very minor rewording of debugging text in manualroute to say "list of
5640 hosts" instead of "hostlist".
5642 2. If verify=header_syntax was set, and a header line with an unqualified
5643 address (no domain) and a large number of spaces between the end of the
5644 name and the colon was received, the reception process suffered a buffer
5645 overflow, and (when I tested it) crashed. This was caused by some obsolete
5646 code that should have been removed. The fix is to remove it!
5648 3. When running in the test harness, delay a bit after writing a bounce
5649 message to get a bit more predictability in the log output.
5651 4. Added a call to search_tidyup() just before forking a reception process. In
5652 theory, someone could use a lookup in the expansion of smtp_accept_max_
5653 per_host which, without the tidyup, could leave open a database connection.
5655 5. Added the variables $recipient_data and $sender_data which get set from a
5656 lookup success in an ACL "recipients" or "senders" condition, or a router
5657 "senders" option, similar to $domain_data and $local_part_data.
5659 6. Moved the writing of debug_print from before to after the "senders" test
5662 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5663 problems for message scanning, either using a data ACL, or using
5664 local_scan() because the Received: header was not generated till after they
5665 were called (in order to set the time as the time of reception completion).
5666 I have revised the way this works. The header is now generated after the
5667 body is received, but before the ACL or local_scan() are called. After they
5668 are run, the timestamp in the header is updated.
5674 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5675 before starting a queue runner without re-exec. This happened only when
5676 deliver_drop_privilege was set or when the Exim user was set to root. The
5677 effect of the bug was that timeouts during subsequent deliveries caused
5678 crashes instead of being properly handled. The handler is now left at its
5679 default (and expected) setting.
5681 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5682 message, again when deliver_drop_privilege is set or Exim is run as root.
5683 The bug described in (1) was not present in this case, but the tidying up
5684 of the other signals was missing. I have made the two cases consistent.
5686 3. The ignore_target_hosts setting on a manualroute router was being ignored
5687 for hosts that were looked up using the /MX notation.
5689 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5692 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5693 operated on the sender address. After changing the $sender_address to <>
5694 for the sender address verify, Exim was re-instated it as the original
5695 (before rewriting) address, but remembering that it had rewritten it, so it
5696 wasn't rewriting it again. This bug also had the effect of breaking the
5697 sender address verification caching when the sender address was rewritten.
5699 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5700 This has been changed so that if the ip literal address matches
5701 ignore_target_hosts, the router declines.
5703 7. Added expansion conditions match_domain, match_address, and match_local_
5704 part (NOT match_host).
5706 8. The placeholder for the Received: header didn't have a length field set.
5708 9. Added code to Exim itself and to exim_lock to test for a specific race
5709 condition that could lead to file corruption when using MBX delivery. The
5710 issue is with the lockfile that is created in /tmp. If this file is removed
5711 after a process has opened it but before that process has acquired a lock,
5712 there is the potential for a second process to recreate the file and also
5713 acquire a lock. This could lead to two Exim processes writing to the file
5714 at the same time. The added code performs the same test as UW imapd; it
5715 checks after acquiring the lock that its file descriptor still refers to
5716 the same named file.
5718 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5719 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5722 11. Added the smtp_active_hostname option. If used, this will typically be made
5723 to depend on the incoming interface address. Because $interface_address is
5724 not set up until the daemon has forked a reception process, error responses
5725 that can happen earlier (such as "too many connections") no longer contain
5728 12. If an expansion in a condition on a "warn" statement fails because a lookup
5729 defers, the "warn" statement is abandoned, and the next ACL statement is
5730 processed. Previously this caused the whole ACL to be aborted.
5732 13. Added the iplsearch lookup type.
5734 14. Added ident_timeout as a log selector.
5736 15. Added tls_certificate_verified as a log selector.
5738 16. Added a global option tls_require_ciphers (compare the smtp transport
5739 option of the same name). This controls incoming TLS connections.
5741 17. I finally figured out how to make tls_require_ciphers do a similar thing
5742 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5743 before starting the TLS session.
5745 18. Tabs are now shown as \t in -bP output.
5747 19. If the log selector return_path_on_delivery was set, Exim crashed when
5748 bouncing a message because it had too many Received: header lines.
5750 20. If two routers both had headers_remove settings, and the first one included
5751 a superfluous trailing colon, the final name in the first list and the
5752 first name in the second list were incorrectly joined into one item (with a
5753 colon in the middle).
5759 1. Added -C and -D options to the exinext utility, mainly to make it easier
5760 to include in the automated testing, but these could be helpful when
5761 multiple configurations are in use.
5763 2. The exinext utility was not formatting the output nicely when there was
5764 an alternate port involved in the retry record key, nor when there was a
5765 message id as well (for retries that were specific to a specific message
5766 and a specific host). It was also confused by IPv6 addresses, because of
5767 the additional colons they contain. I have fixed the IPv4 problem, and
5768 patched it up to do a reasonable job for IPv6.
5770 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5771 delivery, the log line now contains "pipelined" if PIPELINING was used.
5773 4. An SMTP transport process used to panic and die if the bind() call to set
5774 an explicit outgoing interface failed. This has been changed; it is now
5775 treated in the same way as a connect() failure.
5777 5. A reference to $sender_host_name in the part of a conditional expansion
5778 that was being skipped was still causing a DNS lookup. This no longer
5781 6. The def: expansion condition was not recognizing references to header lines
5782 that used bh_ and bheader_.
5784 7. Added the _cache feature to named lists.
5786 8. The code for checking quota_filecount in the appendfile transport was
5787 allowing one more file than it should have been.
5789 9. For compatibility with Sendmail, the command line option
5797 and sets the incoming protocol and host name (for trusted callers). The
5798 host name and its colon can be omitted when only the protocol is to be set.
5799 Note the Exim already has two private options, -pd and -ps, that refer to
5800 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5801 "s", but I don't think that's a major issue.
5803 10. A number of refactoring changes to the code, none of which should affect
5806 (a) The number of logging options was getting close to filling up the
5807 32-bit word that was used as a bit map. I have split them into two classes:
5808 those that are passed in the argument to log_write(), and those that are
5809 only ever tested independently outside of that function. These are now in
5810 separate 32-bit words, so there is plenty of room for expansion again.
5811 There is no change in the user interface or the logging behaviour.
5813 (b) When building, for example, log lines, the code previously used a
5814 macro that called string_cat() twice, in order to add two strings. This is
5815 not really sufficiently general. Furthermore, there was one instance where
5816 it was actually wrong because one of the argument was used twice, and in
5817 one call a function was used. (As it happened, calling the function twice
5818 did not affect the overall behaviour.) The macro has been replaced by a
5819 function that can join an arbitrary number of extra strings onto a growing
5822 (c) The code for expansion conditions now uses a table and a binary chop
5823 instead of a serial search (which was left over from when there were very
5824 few conditions). Also, it now recognizes conditions like "pam" even when
5825 the relevant support is not compiled in: a suitably worded error message is
5826 given if an attempt is made to use such a condition.
5828 11. Added ${time_interval:xxxxx}.
5830 12. A bug was causing one of the ddress fields not to be passed back correctly
5831 from remote delivery subprocesses. The field in question was not being
5832 subsequently used, so this caused to problems in practice.
5834 13. Added new log selectors queue_time and deliver_time.
5836 14. Might have fixed a bug in maildirsizefile handling that threw up
5837 "unexpected character" debug warnings, and recalculated the data
5838 unnecessarily. In any case, I expanded the warning message to give more
5841 15. Added the message "Restricted characters in address" to the statements in
5842 the default ACL that block characters like @ and % in local parts.
5844 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5845 Three changes have been made:
5847 (a) There was a serious bug; a negative response to MAIL caused the whole
5848 recipient domain to be cached as invalid, thereby blocking all messages
5849 to all local parts at the same domain, from all senders. This bug has
5850 been fixed. The domain is no longer cached after a negative response to
5851 MAIL if the sender used is not empty.
5853 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5856 (c) A new callout option, "use_sender" has been added for people who want
5857 the modified behaviour.
5863 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5866 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5867 indeed breaks things for older releases.
5869 3. Added additional logging to the case where there is a problem reading data
5870 from a filter that is running in a subprocess using a pipe, in order to
5871 try to track down a specific problem.
5873 4. Testing facility fudge: when running in the test harness and attempting
5874 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5875 getting "No route to host". Convert this to a timeout.
5877 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5880 6. Some OS don't have socklen_t but use size_t instead. This affects the
5881 fifth argument of getsockopt() amongst other things. This is now
5882 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5883 can be set for individual OS. I have set it for SunOS5, OSF1, and
5884 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5885 some earlier ones do not.
5887 7. Change 4.30/15 was not doing the test caselessly.
5889 8. The standard form for an IPv6 address literal was being rejected by address
5890 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5891 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5892 this, as well as the form without the "IPv6" on the front (but only when
5893 address literals are enabled, of course).
5895 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5897 10. Exim crashed if a message with an empty sender address specified by -f
5898 encountered a router with an errors_to setting. This could be provoked only
5899 by a command such as
5903 where an empty string was supplied; "<>" did not hit this bug.
5905 11. Installed PCRE release 4.5.
5907 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5908 remained set. It is now erased.
5910 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5911 times from message ids (which are base 36 rather than the normal 62).
5913 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5914 were being counted as actual protocol errors, and logged if the log
5915 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5916 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5917 and DATA following a set of rejected RCPTs do not count as protocol errors.
5918 In other words, Exim assumes they were pipelined, though this may not
5919 actually be the case. Of course, in all cases the client gets an
5920 appropriate error code.
5922 15. If a lookup fails in an ACL condition, a message about the failure may
5923 be available; it is used if testing the ACL cannot continue, because most
5924 such messages specify what the cause of the deferral is. However, some
5925 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5926 that caused an old message to be retained and used if a later statement
5927 caused a defer, replacing the real cause of the deferral.
5929 16. If an IP address had so many PTR records that the DNS lookup buffer
5930 was not large enough to hold them, Exim could crash while trying to process
5931 the truncated data. It now detects and logs this case.
5933 17. Further to 4.21/58, another change has been made: if (and only if) the
5934 first line of a message (the first header line) ends with CRLF, a bare LF
5935 in a subsequent header line has a space inserted after it, so as not to
5936 terminate the header.
5938 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5939 unnecessarily, used atoi() instead of strtol(), and didn't check the
5940 termination when getting file sizes from file names by regex.
5942 19. Completely re-implemented the support for maildirsize files, in the light
5943 of a number of problems with the previous contributed implementation
5944 (4.30/29). In particular:
5946 . If the quota is zero, the maildirsize file is maintained, but no quota is
5949 . If the maildir directory does not exist, it is created before any attempt
5950 to write a maildirsize file.
5952 . The quota value in the file is just a cache; if the quota is changed in
5953 the transport, the new value overrides.
5955 . A regular expression is available for excluding directories from the
5958 20. The autoreply transport checks the characters in options that define the
5959 message's headers; it allows continued headers, but it was checking with
5960 isspace() after an embedded newline instead of explicitly looking for a
5963 21. If all the "regular" hosts to which an address was routed had passed their
5964 expiry times, and had not reached their retry times, the address was
5965 bounced, even if fallback hosts were defined. Now Exim should go on to try
5968 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5969 equivalent code in the SMTP transport. Some hosts send humungous responses
5970 to HELO/EHLO, more than 1024 it seems.
5972 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5973 gives compiler warnings in some environments. I've now done it "properly",
5976 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5977 (because of problems with the built-in one) was declared to return uschar *
5978 instead of char *, causing compiler failure.
5980 25. Fixed a file descriptor leak when processing alias/forward files.
5982 26. Fixed a minor format string issue in dbfn.c.
5984 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5986 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5987 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5990 29. When a sender address is verified, it is cached, to save repeating the test
5991 when there is more than one recipient in a message. However, when the
5992 verification involves a callout, it is possible for different callout
5993 options to be set for different recipients. It is too complicated to keep
5994 track of this in the cache, so now Exim always runs a verification when a
5995 callout is required, relying on the callout cache for the optimization.
5996 The overhead is duplication of the address routing, but this should not be
5999 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6000 to be verified with callout=postmaster, and the main callout worked but the
6001 postmaster check failed, the verification correctly failed. However, if a
6002 subsequent RCPT command asked for sender verification *without* the
6003 postmaster check, incorrect caching caused this verification also to fail,
6006 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6007 it was not caching the DNS options (qualify_single, search_parents) that
6008 were used when the lookup failed. A subsequent lookup with different
6009 options therefore always gave the same answer, though there were cases
6010 where it should not have. (Example: a "domains = !$mx_any" option on a
6011 dnslookup router: the "domains" option is always processed without any
6012 widening, but the router might have qualify_single set.) Now Exim uses the
6013 cached value only when the same options are set.
6015 32. Added John Jetmore's "exipick" utility to the distribution.
6017 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6018 than a timeout (e.g. a certificate is required, and is not provided), an
6019 Exim server now closes the connection immediately. Previously it waited for
6020 the client to close - but if the client is SSL, it seems that they each
6021 wait for each other, leading to a delay before one of them times out.
6023 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6024 maintained 0.8.x compatibility because I don't think many are using it, and
6025 it is clearly obsolete.
6027 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6030 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6031 client certificate was expired. A simple patch fixes this, though I don't
6032 understand the full logic of why the verify callback is called multiple
6035 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6036 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6037 which causes problems with some clients (such as the Certicom SSL Plus
6038 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6039 disables the coutermeasure allowing Eudora to connect."
6041 38. Exim was not checking that a write() to a log file succeeded. This could
6042 lead to Bad Things if a log got too big, in particular if it hit a file
6043 size limit. Exim now panics and dies if it cannot write to a log file, just
6044 as it does if it cannot open a log file.
6046 39. Modified OS/Makefile-Linux so that it now contains
6048 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6050 The two -D definitions ensure that Exim is compiled with large file
6051 support, which makes it possible to handle log files that are bigger than
6054 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6055 instance) a domain was checked against a named list that involved a lookup,
6056 causing $domain_data to be set, then another domain was checked against the
6057 same list, then the first domain was re-checked, the value of $domain_data
6058 after the final check could be wrong. In particular, if the second check
6059 failed, it could be set empty. This bug probably also applied to
6062 41. The strip_trailing_dot option was not being applied to the address given
6063 with the -f command-line option.
6065 42. The code for reading a message's header from the spool was incrementing
6066 $received_count, but never initializing it. This meant that the value was
6067 incorrect (doubled) while delivering a message in the same process in which
6068 it was received. In the most common configuration of Exim, this never
6069 happens - a fresh exec is done - but it can happen when
6070 deliver_drop_privilege is set.
6072 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6073 it now includes up to 150 characters of the unexpected data in the log
6076 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6077 and building data strings. The size of both of these buffers was 10 000
6078 bytes - far larger than anybody would *ever* want, thought I. Needless to
6079 say, somebody hit the limit. I have increased the maximum line length to
6080 20 000 and the maximum data length of concatenated lines to 100 000. I have
6081 also fixed two bugs, because there was no checking on these buffers. Tsk,
6082 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6083 buffer is too small.
6085 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6086 lsearch lookups. Now it does.
6088 46. When parsing a route_list item in a manualroute router, a fixed-length
6089 buffer was used for the list of hosts. I made this 1024 bytes long,
6090 thinking that nobody would ever have a list of hosts that long. Wrong.
6091 Somebody had a whole pile of complicated expansion conditions, and the
6092 string was silently truncated, leading to an expansion error. It turns out
6093 that it is easier to change to an unlimited length (owing to other changes
6094 that have happened since this code was originally written) than to build
6095 structure for giving a limitation error. The length of the item that
6096 expands into the list of hosts is now unlimited.
6098 47. The lsearch lookup could not handle data where the length of text line was
6099 more than 4095 characters. Such lines were truncated, leading to shortened
6100 data being returned. It should now handle lines of any length.
6102 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6103 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6106 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6107 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6108 no longer happens in comment lines. A list of replacements is now placed
6109 at the head of all of the source files, except those whose only change is
6110 to replace PERL_COMMAND in the very first #! line.
6112 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6113 messages on the queue, with a bottom-up merge sort, using code contributed
6114 by Michael Haardt. This should make operations like -bp somewhat faster on
6115 large queues. It won't affect queue runners, except when queue_run_in_order
6118 51. Installed eximstats 1.31 in the distribution.
6120 52. Added support for SRV lookups to the dnslookup router.
6122 53. If an ACL referred to $message_body or $message_body_end, the value was not
6123 reset for any messages that followed in the same SMTP session.
6125 54. The store-handling optimization for building very long strings was not
6126 differentiating between the different store pools. I don't think this
6127 actually made any difference in practice, but I've tidied it.
6129 55. While running the routers to verify a sender address, $sender_address
6130 was still set to the sender address. This is wrong, because when routing to
6131 send a bounce to the sender, it would be empty. Therefore, I have changed
6132 it so that, while verifying a sender address, $sender_address is set to <>.
6133 (There is no change to what happens when verifying a recipient address.)
6135 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6136 target A or AAAA records (if not already returned) without resetting the
6137 qualify_single or search_parents options of the DNS resolver. These are
6138 inappropriate in this case because the targets of MX and SRV records must
6139 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6140 target that, when qualified, matched something in the local domain. These
6141 two options are now turned off when doing these lookups.
6143 57. It seems that at least some releases of Reiserfs (which does not have the
6144 concept of a fixed number of inodes) returns zero and not -1 for the
6145 number of available inodes. This interacted badly with check_spool_inodes,
6146 which assumed that -1 was the "no such thing" setting. What I have done is
6147 to check that the total number of inodes is greater than zero before doing
6148 the test of how many are available.
6150 58. When a "warn" ACL statement has a log_message modifier, the message is
6151 remembered, and not repeated. This is to avoid a lot of repetition when a
6152 message has many recipients that cause the same warning to be written.
6153 However, Exim was preserving the list of already written lines for an
6154 entire SMTP session, which doesn't seem right. The memory is now reset if a
6155 new message is started.
6157 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6158 debugging output unless log_rewrite was also set.
6160 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6161 of (int)(handle) when we know that handle contains (void *)(-1).
6163 61. The Exim daemon panic-logs an error return when it closes the incoming
6164 connection. However "connection reset by peer" seems to be common, and
6165 isn't really an error worthy of noting specially, so that particular error
6168 62. When Exim is trying to find all the local interfaces, it used to panic and
6169 die if the ioctl to get the interface flags failed. However, it seems that
6170 on at least one OS (Solaris 9) it is possible to have an interface that is
6171 included in the list of interfaces, but for which you get a failure error
6172 for this call. This happens when the interface is not "plumbed" into a
6173 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6174 failure of the "get flags" call assumes that the interface is down.
6176 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6177 makes life easier for people who are doing arithmetic on fields extracted
6178 from dates, where you often get leading zeros that should not be
6179 interpreted as octal.
6181 64. Added qualify_domain to the redirect router, to override the global
6184 65. If a pathologically long header line contained very many addresses (the
6185 report of this problem mentioned 10 000) and each of them was rewritten,
6186 Exim could use up a very large amount of memory. (It kept on making new
6187 copies of the header line as it rewrote, and never released the old ones.)
6188 At the expense of a bit more processing, the header rewriting function has
6189 been changed so that it no longer eats memory in this way.
6191 66. The generation of the Received: header has been moved from the time that a
6192 message starts to be received, to the time that it finishes. The timestamp
6193 in the Received: header should now be very close to that of the <= log
6194 line. There are two side-effects of this change:
6196 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6197 logged header lines no longer include the local Received: line, because
6198 it has not yet been created. The same applies to a copy of the message
6199 that is returned to a non-SMTP sender when a message is rejected.
6201 (b) When a filter file is tested using -bf, no additional Received: header
6202 is added to the test message. After some thought, I decided that this
6205 This change does not affect the value of $received_for. It is still set
6206 after address rewriting, but before local_scan() is called.
6208 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6210 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6211 gave an unhelpful panic error message, and a defer error. I have managed to
6212 change this behaviour so that it now rejects any supplied certificate,
6213 which seems right, as the list of acceptable certificates is empty.
6215 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6216 gave an unhelpful defer error. I have not managed to make this reject any
6217 supplied certificates, but the error message it gives is "no certificate
6218 supplied", which is not helpful.
6220 70. exigrep's output now also includes lines that are not associated with any
6221 message, but which match the given pattern. Implemented by a patch from
6222 Martin Sluka, which also tidied up the Perl a bit.
6224 71. Recipient callout verification, like sender verification, was using <> in
6225 the MAIL FROM command. This isn't really the right thing, since the actual
6226 sender may affect whether the remote host accepts the recipient or not. I
6227 have changed it to use the actual sender in the callout; this means that
6228 the cache record is now keyed on a recipient/sender pair, not just the
6229 recipient address. There doesn't seem to be a real danger of callout loops,
6230 since a callout by the remote host to check the sender would use <>.
6231 [SEE ABOVE: changed after hitting problems.]
6233 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6234 temporary errors. However, in the case of such a code being given after
6235 the end of a data transmission (i.e. after ".") Exim was failing to write
6236 a retry record for the message. (Yes, there was some broken host that was
6237 actually sending 8xx at this point.)
6239 73. An unknown lookup type in a host list could cause Exim to panic-die when
6240 the list was checked. (An example that provoked this was putting <; in the
6241 middle of a list instead of at the start.) If this happened during a DATA
6242 ACL check, a -D file could be left lying around. This kind of configuration
6243 error no longer causes Exim to die; instead it causes a defer error. The
6244 incident is still logged to the main and panic logs.
6246 74. Buglet left over from Exim 3 conversion. The message "too many messages
6247 in one connection" was written to the rejectlog but not the mainlog, except
6248 when address rewriting (yes!) was being logged.
6250 75. Added write_rejectlog option.
6252 76. When a system filter was run not as root (that is, when system_filter_user
6253 was set), the values of the $n variables were not being returned to the
6254 main process; thus, they were not subsequently available in the $sn
6257 77. Added +return_path_on_delivery log selector.
6259 78. A connection timeout was being treated differently from recipients deferred
6260 when testing hosts_max_try with a message that was older than the host's
6261 retry timeout. (The host should not be counted, thus allowing all hosts to
6262 be tried at least once before bouncing.) This may have been the cause of an
6263 occasionally reported bug whereby a message would remain on the queue
6264 longer than the retry timeout, but would be bounced if a delivery was
6265 forced. I say "may" because I never totally pinned down the problem;
6266 setting up timeout/retry tests is difficult. See also the next item.
6268 79. The ultimate address timeout was not being applied to errors that involved
6269 a combination of host plus message (for example, a timeout on a MAIL
6270 command). When an address resolved to a number of possible hosts, and they
6271 were not all tried for each delivery (e.g. because of hosts_max_try), a
6272 message could remain on the queue longer than the retry timeout.
6274 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6277 81. Fixed an obscure SMTP outgoing bug which required at least the following
6278 conditions: (a) there was another message waiting for the same server;
6279 (b) the server returned 5xx to all RCPT commands in the first message so
6280 that the message was not completed; (c) the server dropped the connection
6281 or gave a negative response to the RSET that Exim sends to abort the
6282 transaction. The observed case was a dropped connection after DATA that had
6283 been sent in pipelining mode. That is, the server had advertised PIPELINING
6284 but was not implementing it correctly. The effect of the bug was incorrect
6285 behaviour, such as trying another host, and this could lead to a crash.
6291 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6292 and daemon.c were passed as pointers to ints; they should have been
6293 pointers to socklen_t variables (which are typically unsigned ints).
6295 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6298 3. Fixed a really odd bug that affected only the testing scheme; patching a
6299 certain fixed string in the binary changed the value of another string that
6300 happened to be identical to the end of the original first string.
6302 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6303 name", it returns that address as the IP address. On some operating
6304 systems (e.g. Solaris), it also passes back the IP address string as the
6305 "host name". However, on others (e.g. Linux), it passes back an empty
6306 string. Exim wasn't checking for this, and was changing the host name to an
6307 empty string, assuming it had been canonicalized.
6309 5. Although rare, it is permitted to have more than one PTR record for a given
6310 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6311 all the names associated with an address, because they do in Solaris.
6312 However, it seems that they do not in Linux for data that comes from the
6313 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6314 I found this out when I moved to a new Linux workstation and tried to run
6315 the Exim test suite.
6317 To get round this problem I have changed the code so that it now does its
6318 own call to the DNS to look up PTR records when searching for a host name.
6319 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6320 addresses that are only in /etc/hosts are still found.
6322 This behaviour is, however, controlled by an option called host_lookup_
6323 order, which defaults to "bydns:byaddr". If people want to use the other
6324 order, or indeed, just use one or the other means of lookup, they can
6325 specify it in this variable.
6327 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6328 some operating systems, this comes back from gethostbyaddr() as an empty
6329 string, and this is what Exim used to test for. However, it seems that in
6330 other systems, "." is yielded. Exim now tests for this case too.
6332 7. The values of check_spool_space and check_log_space are now held internally
6333 as a number of kilobytes instead of an absolute number of bytes. If a
6334 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6335 kilobyte. This means that much larger values can be stored.
6337 8. Exim monitor: an attempt to get the action menu when not actually pointing
6338 at a message produces an empty menu entitled "No message selected". This
6339 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6340 no entries in it ("Shell widget menu has zero width and/or height"). So I
6341 have added a single, blank menu entry in this case.
6343 9. Added ${quote_local_part.
6345 10. MIME decoding is now applied to the contents of Subject: header lines when
6348 11. Now that a reference to $sender_host_address automatically causes a reverse
6349 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6350 host lookup before query-style lookups in lists that might use this
6351 variable. This has therefore been abolished, and the "net-" prefix is no
6352 longer necessary for query-style lookups.
6354 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6355 have been a typo for LFLAGS, so it has been changed.
6357 13. The install script calls Exim with "-C /dev/null" in order to find the
6358 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6359 to be output. However, since Exim outputs its version number before the
6360 error, it didn't break the script. It just looked ugly. I fixed this by
6361 always allowing "-C /dev/null" if the caller is root.
6363 14. Ignore overlarge ACL variable number when reading spool file - insurance
6364 against a later release with more variables having written the file.
6366 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6367 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6368 this, as well as the form without the "IPv6" on the front.
6370 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6371 OS/Makefile-Darwin file.
6373 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6374 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6376 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6377 variable. However, some OS use ints for inodes. I've added cast to long int
6378 to get rid of the compiler warning.
6380 19. I had forgotten to lock out "/../" in configuration file names when
6381 ALT_CONFIG_PREFIX was set.
6383 20. Routers used for verification do not need to specify transports. However,
6384 if such a router generated a host list, and callout was configured, Exim
6385 crashed, because it could not find a port number from the (non-existent)
6386 transport. It now assumes port 25 in this circumstance.
6388 21. Added the -t option to exigrep.
6390 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6391 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6392 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6393 them is set, LOOKUP_LSEARCH is forced.
6395 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6397 24. Added sender and host information to the "rejected by local_scan()" log
6398 line; previously there was no indication of these.
6400 25. Added .include_if_exists.
6402 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6403 Linux. It turns out that not all file systems support this. Apparently some
6404 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6405 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6406 not supported on the file descriptor, is now ignored when Exim is trying to
6407 sync a directory. This applies only to Linux.
6409 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6411 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6412 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6415 29. Added support for maildirsize files from supplied patch (modified a bit).
6417 30. The use of :fail: followed by an empty string could lead Exim to respond to
6418 sender verification failures with (e.g.):
6420 550 Verification failed for <xxx>
6421 550 Sender verify failed
6423 where the first response line was missing the '-' that indicates it is not
6424 the final line of the response.
6426 31. The loop for finding the name of the user that called Exim had a hardwired
6427 limit of 10; it now uses the value of finduser_retries, which is used for
6428 all other user lookups.
6430 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6433 33. Exim was neglecting to zero errno before one call of strtol() when
6434 expanding a string and expecting an integer value. On some systems this
6435 resulted in spurious "integer overflow" errors. Also, it was casting the
6436 result into an int without checking.
6438 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6439 did not work. The code looks as if it has *never* worked, though it appears
6440 to have been documented since at least release 1.62. I have made it work.
6442 35. The "timeout_DNS" error in retry rules, also documented since at least
6443 1.62, also never worked. As it isn't clear exactly what this means, and
6444 clearly it isn't a major issue, I have abolished the feature by treating it
6445 as "timeout", and writing a warning to the main and panic logs.
6447 36. The display of retry rules for -brt wasn't always showing the error code
6450 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6451 timeout_connect_A, timeout_connect_MX.
6453 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6454 to the empty sender.
6456 39. The daemon was not analysing the content of -oX till after it had closed
6457 stderr and disconnected from the controlling terminal. This meant that any
6458 syntax errors were only noted on the panic log, and the return code from
6459 the command was 0. By re-arranging the code a little, I've made the
6460 decoding happen first, so such errors now appear on stderr, and the return
6461 code is 1. However, the actual setting up of the sockets still happens in
6462 the disconnected process, so errors there are still only recorded on the
6465 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6466 connections (as happens on some IP stacks) was logged at start up time as
6467 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6468 it from "IPv6 and IPv4", which means that two separate sockets are being
6471 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6472 says whether AF_INET or AF_INET6 was passed as an argument.
6474 42. Exiwhat output was messed up when time zones were included in log
6477 43. Exiwhat now gives more information about the daemon's listening ports,
6478 and whether -tls-on-connect was used.
6480 44. The "port" option of the smtp transport is now expanded.
6482 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6483 silently ignored. Now an error message is written to the main and panic
6486 46. There's a new ACL modifier called "logwrite" which writes to a log file
6487 as soon as it is encountered.
6489 47. Added $local_user_uid and $local_user_gid at routing time.
6491 48. Exim crashed when trying to verify a sender address that was being
6494 49. Exim was recognizing only a space character after ".include". It now also
6495 recognizes a tab character.
6497 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6498 extracting the relevant information from the specification. The man page no
6499 longer contains scrambled data for the -d option, and I've added a section
6500 at the front about calling Exim under different names.
6502 51. Added "extra_headers" argument to the "mail" command in filter files.
6504 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6507 53. Installed eximstats 1.29.
6509 54. Added transport_filter_timeout as a generic transport option.
6511 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6512 Cc: header lines. This was required by RFC 822, but it not required by RFC
6515 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6516 incoming messages that did not have them. Now it does so only if the
6517 message originates locally, that is, if there is no associated remote host
6518 address. When Resent- header lines are present, this applies to the Resent-
6519 lines rather than the non-Resent- lines.
6521 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6522 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6524 58. Messages for configuration errors now include the name of the main
6525 configuration file - useful now that there may be more than one file in a
6526 list (.included file names were always shown).
6528 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6529 for those rare installations that do not start the daemon as root or run it
6530 setuid root. I've cut out the call to initgroups() if the daemon is not
6533 60. The Exim user and group can now be bound into the binary as text strings
6534 that are looked up at the start of Exim's processing.
6536 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6538 62. Added $mailstore_basename variable.
6540 63. Installed patch to sieve.c from Michael Haardt.
6542 64. When Exim failed to open the panic log after failing to open the main log,
6543 the original message it was trying to log was written to stderr and debug
6544 output, but if they were not available (the usual case in production), it
6545 was lost. Now it is written to syslog before the two lines that record the
6546 failures to open the logs.
6548 65. Users' Exim filters run in subprocesses under the user's uid. It is
6549 possible for a "deliver" command or an alias in a "personal" command to
6550 provoke an address rewrite. If logging of address rewriting is configured,
6551 this fails because the process is not running as root or exim. There may be
6552 a better way of dealing with this, but for the moment (because 4.30 needs
6553 to be released), I have disabled address rewrite logging when running a
6554 filter in a non-root, non-exim process.
6560 1. The buildconfig auxiliary program wasn't quoting the value set for
6561 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6562 not defined. This bug was masked in 4.22 by the effect that was fixed in
6565 2. Some messages that were rejected after a message id was allocated were
6566 shown as "incomplete" by exigrep. It no longer does this for messages that
6567 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6569 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6570 have allow_domain_literals set, the ID did not get logged in the <= line.
6571 Domain literals are now always recognized in Message-ID: header lines.
6573 4. The first argument for a ${extract expansion item is the key name or field
6574 number. Leading and trailing spaces in this item were not being ignored,
6575 causing some misleading effects.
6577 5. When deliver_drop_privilege was set, single queue runner processes started
6578 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6579 same command in the process it spins off) were not dropping privilege.
6581 6. When the daemon running as "exim" started a queue runner, it always
6582 re-executed Exim in the spun-off process. This is a waste of effort when
6583 deliver_drop_privilege is set. The new process now just calls the
6584 queue-runner function directly.
6590 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6593 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6594 ignored. Though the use of -oP was forcing the writing of a pid file, it
6595 was always written to the default place.
6597 3. If the message "no IP address found for host xxxx" is generated during
6598 incoming verification, it is now followed by identification of the incoming
6599 connection (so you can more easily find what provoked it).
6601 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6603 5. Added some features to "harden" Exim a bit more against certain attacks:
6605 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6606 be put in Local/Makefile. This is like the never_users runtime option,
6607 but it cannot be overridden. The default setting is "root".
6609 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6610 prefix string with which any file named in a -C command line option
6613 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6614 is retained for -C and -D only if the caller of Exim is root. Without
6615 it, the exim user may also use -C and -D and retain privilege.
6617 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6618 command line option is disabled.
6620 6. Macro names set by the -D option must start with an upper case letter, just
6621 like macro names defined in the configuration file.
6623 7. Added "dereference=" facility to LDAP.
6625 8. Two instances of the typo "uknown" in the source files are fixed.
6627 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6628 the Configure-Makefile script screwed up while processing it.
6630 10. Incorporated PCRE 4.4.
6632 11. The SMTP synchronization check was not operating right at the start of an
6633 SMTP session. For example, it could not catch a HELO sent before the client
6634 waited for the greeting. There is now a check for outstanding input at the
6635 point when the greeting is written. Because of the duplex, asynchronous
6636 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6637 way, but not yet received, when the check is performed.
6639 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6640 on TCP/IP sockets, because this apparently causes some broken clients to
6643 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6644 unchanged) from the Cygwin maintainer.
6646 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6647 was supported instead of testing for mbx. Effectively a typo.
6649 15. The spa authenticator server code was not checking that the input it
6650 received was valid base64.
6652 16. The debug output line for the "set" modifier in ACLs was not showing the
6653 name of the variable that was being set.
6655 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6657 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6658 lookup on its own. Something else was needed to trigger the lookup. For
6659 example, a match in host_lookup or the need for a host name in a host list.
6660 Now, if $sender_host_name is referenced and the host name has not yet been
6661 looked up, a lookup is performed. If the lookup fails, the variable remains
6662 empty, and $host_lookup_failed is set to "1".
6664 19. Added "eqi" as a case-independent comparison operator.
6666 20. The saslauthd authentication condition could segfault if neither service
6667 nor realm was specified.
6669 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6670 error message that was logged was misleading, and incoming SMTP
6671 connections were dropped. The message is now more accurate, and temporary
6672 errors are given to SMTP connections.
6674 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6675 (except RSET) with a 421 error, until QUIT is received. However, it was
6676 failing to send a response to QUIT.
6678 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6679 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6680 ACL is now run after the helo_try_verify_hosts code.
6682 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6683 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6684 case-independent, so other case variants are also recognized). Apparently
6685 some systems use these upper case variants.
6687 25. If more than two messages were waiting for the same host, and a transport
6688 filter was specified for the transport, Exim sent two messages over the
6689 same TCP/IP connection, and then failed with "socket operation on non-
6690 socket" when it tried to send the third.
6692 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6694 27. The extern definition of crypt16() in expand.c was not being excluded when
6695 the OS had its own crypt16() function.
6697 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6698 as a preferred synonym for return_size_limit, both as an option and as an
6701 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6703 30. Changed the default configuration ACL to relax the local part checking rule
6704 for addresses that are not in any local domains. For these addresses,
6705 slashes and pipe symbols are allowed within local parts, but the sequence
6706 /../ is explicitly forbidden.
6708 31. SPA server authentication was not clearing the challenge buffer before
6711 32. log_message in a "warn" ACL statement was writing to the reject log as
6712 well as to the main log, which contradicts the documentation and doesn't
6713 seem right (because no rejection is happening). So I have stopped it.
6715 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6716 However, I am unable to do any testing of this.
6718 34. Fixed an infelicity in the appendfile transport. When checking directories
6719 for a mailbox, to see if any needed to be created, it was accidentally
6720 using path names with one or more superfluous leading slashes; tracing
6721 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6723 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6724 contents are added to the log line that is written for every discarded
6725 recipient. (Previously a log_message setting was ignored.)
6727 36. The ${quote: operator now quotes the string if it is empty.
6729 37. The install script runs exim in order to find its version number. If for
6730 some reason other than non-existence or emptiness, which it checks, it
6731 could not run './exim', it was installing it with an empty version number,
6732 i.e. as "exim-". This error state is now caught, and the installation is
6735 38. An argument was missing from the function that creates an error message
6736 when Exim fails to connect to the socket for saslauthd authentication.
6737 This could cause Exim to crash, or give a corrupted message.
6739 39. Added isip, isip4, and isip6 to ${if conditions.
6741 40. The ACL variables $acl_xx are now saved with the message, and can be
6742 accessed later in routers, transports, and filters.
6744 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6745 strings in the file are not string-expanded.
6747 42. If a MAIL command specified a SIZE value that was too large to fit into an
6748 int variable, the check against message_size_limit failed. Such values are
6749 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6750 day this will have to be increased, but I don't think I want to be around
6751 when emails are that large.
6758 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6759 iconv() is not standard in FreeBSD.
6761 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6762 IPv6 enabled. The observed symptom was a segmentation fault on return from
6763 the function os_common_find_running_interfaces() in src/os.c.
6765 3. In the check_special_case() function in daemon.c I had used "errno" as an
6766 argument name, which causes warnings on some systems. This was basically a
6767 typo, since it was named "eno" in the comments!
6769 4. The code that waits for the clock to tick (at a resolution of some fraction
6770 of a second) so as to ensure message-id uniqueness was always waiting for
6771 at least one whole tick, when it could have waited for less. [This is
6772 almost certainly not relevant at current processor speeds, where it is
6773 unlikely to ever wait at all. But we try to future-proof.]
6775 5. The function that sleeps for a time interval that includes fractions of a
6776 second contained a race. It did not block SIGALRM between setting the
6777 timer, and suspending (a couple of lines later). If the interval was short
6778 and the sigsuspend() was delayed until after it had expired, the suspension
6779 never ended. On busy systems this could lead to processes getting stuck for
6782 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6783 process, before it forks any delivery processes. The open lookup caching
6784 mechanism meant that the open file or database connection was passed into
6785 the delivery process. The problem was that delivery processes always tidy
6786 up cached lookup data. This could cause a problem for the next delivery
6787 process started by the queue runner, because the external queue runner
6788 process does not know about the closure. So the next delivery process
6789 still has data in the lookup cache. In the case of a file lookup, there was
6790 no problem because closing a file descriptor in a subprocess doesn't affect
6791 the parent. However, if the lookup was caching a connection to a database,
6792 the connection was closed, and the second delivery process was likely to
6793 see errors such as "PGSQL: query failed: server closed the connection
6794 unexpectedly". The problem has been fixed by closing all cached lookups
6795 in a queue runner before running a delivery process.
6797 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6798 seem to have the "const" qualifier which it has on other OS. I've
6801 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6802 *of the same type* (client or server) with the same public name that an
6803 error should be diagnosed.
6805 9. When Exim looked up a host name for an IP address, but failed to find the
6806 original IP address when looking up the host name (a safety check), it
6807 output the message "<ip address> does not match any IP for NULL", which was
6808 confusing, to say the least. The bug was that the host name should have
6809 appeared instead of "NULL".
6811 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6812 user that is built into the binary, and the -C or -D options is used, root
6813 privilege is dropped before the configuration file is read. In addition,
6814 logging is switched to stderr instead of the normal log files. If the
6815 configuration then re-defines the Exim user, the unprivileged environment
6816 is probably not what is expected, so Exim logs a panic warning message (but
6819 However, if deliver_drop_privilege is set, the unprivileged state may well
6820 be exactly what is intended, so the warning has been cut out in that case,
6821 and Exim is allowed to try to write to its normal log files.
6827 1. smtp_return_error_details was not giving details for temporary sender
6828 or receiver verification errors.
6830 2. Diagnose a configuration error if two authenticators have the same public
6833 3. Exim used not to create the message log file for a message until the first
6834 delivery attempt. This could be confusing when incoming messages were held
6835 for policy or load reasons. The message log file is now created at the time
6836 the message is received, and an initial "Received" line is written to it.
6838 4. The automatically generated man page for command line options had a minor
6839 bug that caused no ill effects; however, a more serious problem was that
6840 the procedure for building the man page automatically didn't always
6841 operate. Consequently, release 4.20 contains an out-of-date version. This
6842 shouldn't happen again.
6844 5. When building Exim with embedded Perl support, the script that builds the
6845 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6846 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6848 6. The freeze_tell option was not being used for messages that were frozen on
6849 arrival, either by an ACL or by local_scan().
6851 7. Added the smtp_incomplete_transaction log selector.
6853 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6854 was accepting AUTH without a new EHLO.
6856 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6857 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6860 10. Logging of TCP/IP connections (when configured) now happens in the main
6861 daemon process instead of the child process, so that the TCP/IP connection
6862 count is more accurate (but it can never be perfect).
6864 11. The use of "drop" in a nested ACL was not being handled correctly in the
6865 outer ACL. Now, if condition failure induced by the nested "drop" causes
6866 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6867 or "require"), the connection is dropped.
6869 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6870 that yield "discard" can now be used with an "accept" or a "discard" verb,
6871 but an error is generated for any others (because I can't see a useful way
6872 to define what should happen).
6874 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6875 are now processed in the same way as lines in the Exim configuration file.
6876 In particular, continuation lines are supported.
6878 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6880 15. Added -ti meaning -t -i.
6882 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6883 domains, and warn by logging if others are found.
6885 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6886 returned from GIFCONF when Exim is trying to find the list of interfaces on
6887 a host. The code in os.c has been modified to copy each ifreq to an aligned
6888 structure in all cases.
6890 Also, in some cases, the returned ifreq's were being copied to a 'struct
6891 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6892 means the last couple of bytes of an IPv6 address could be chopped if the
6893 ifreq contained only a normal sockaddr (14 bytes storage).
6895 18. Named domain lists were not supported in the hosts_treat_as_local option.
6896 An entry such as +xxxx was not recognized, and was treated as a literal
6899 19. Ensure that header lines added by a DATA ACL are included in the reject log
6900 if the ACL subsequently rejects the message.
6902 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6903 MD5 (which is deprecated).
6905 21. When testing a filter file using -bf, Exim was writing a message when it
6906 took the sender from a "From " line in the message, but it was not doing so
6907 when it took $return_path from a Return-Path: header line. It now does.
6909 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6910 with a valid header line field name (a series of printing characters
6911 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6913 23. Changed "disc" in the source to "disk" to conform to the documentation and
6914 the book and for uniformity.
6916 24. Ignore Sendmail's -Ooption=value command line item.
6918 25. When execve() failed while trying to run a command in a pipe transport,
6919 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6920 could be confused with a return value of 69 from the command itself. This
6921 has been changed to 127, the value the shell returns if it is asked to run
6922 a non-existent command. The wording for the related log line suggests a
6923 non-existent command as the problem.
6925 26. If received_header_text expands to an empty string, do not add a Received:
6926 header line to the message. (Well, it adds a token one on the spool, but
6927 marks it "old" so that it doesn't get used or transmitted.)
6929 27. Installed eximstats 1.28 (addition of -nt option).
6931 28. There was no check for failure on the call to getsockname() in the daemon
6932 code. This can fail if there is a shortage of resources on the system, with
6933 ENOMEM, for example. A temporary error is now given on failure.
6935 29. Contrary to the C standard, it seems that in some environments, the
6936 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6937 program. Exim now does this explicitly; it affects the formatting of
6938 timestamps using strftime().
6940 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6941 complaints. I've now initialized all the variables, to avoid this.
6943 32. Header lines added by a system filter were not being "seen" during
6944 transport-time rewrites.
6946 33. The info_callback() function passed to OpenSSL is set up with type void
6947 (*)(SSL *, int, int), as described somewhere. However, when calling the
6948 function (actually a macro) that sets it up, the type void(*)() is
6949 expected. I've put in a cast to prevent warnings from picky compilers.
6951 34. If a DNS black list lookup found a CNAME record, but there were no A
6952 records associated with the domain it pointed at, Exim crashed.
6954 35. If a DNS black list lookup returned more than one A record, Exim ignored
6955 all but the first. It now scans all returned addresses if a particular IP
6956 value is being sought. In this situation, the contents of the
6957 $dnslist_value variable are a list of all the addresses, separated by a
6960 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6961 to accept a host name and all its aliases if the forward lookup for any of
6962 them yielded the IP address of the incoming connection. Now it accepts only
6963 those names whose forward lookup yields the correct IP address. Any other
6964 names are discarded. This closes a loophole whereby a rogue DNS
6965 administrator could create reverse DNS records to break through a
6966 wildcarded host restriction in an ACL.
6968 37. If a user filter or a system filter that ran in a subprocess used any of
6969 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6970 the wrong values were passed to the pipe command ($thisaddress had the
6971 value of $0, $0 had the value of $1, etc). This bug was introduced by
6972 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6974 38. Improved the line breaking for long SMTP error messages from ACLs.
6975 Previously, if there was no break point between 40 and 75 characters, Exim
6976 left the rest of the message alone. Two changes have been made: (a) I've
6977 reduced the minimum length to 35 characters; (b) if it can't find a break
6978 point between 35 and 75 characters, it looks ahead and uses the first one
6979 that it finds. This may give the occasional overlong line, but at least the
6980 remaining text gets split now.
6982 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6983 file descriptors might be low, and that setting 1000 would always raise it.
6984 It turns out that in some environments, the limit is already over 1000 and
6985 that lowering it causes trouble. So now Exim takes care not to decrease it.
6987 40. When delivering a message, the value of $return_path is set to $sender_
6988 address at the start of routing (routers may change the value). By an
6989 oversight, this default was not being set up when an address was tested by
6990 -bt or -bv, which affected the outcome if any router or filter referred to
6993 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6994 log only when "message" or "log_message" are set. However, if one of the
6995 conditions was an address verification, or a call to a nested ACL, the
6996 messages generated by the underlying test were being passed through. This
6997 no longer happens. The underlying message is available in $acl_verify_
6998 message for both "message" and "log_message" expansions, so it can be
6999 passed through if needed.
7001 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7002 new expansion $bh_ to give the encoded byte string without charset
7003 translation. Translation happens only if iconv() is available; HAVE_ICONV
7004 indicates this at build time. HEADERS_CHARSET gives the charset to
7005 translate to; headers_charset can change it in the configuration, and
7006 "headers charset" can change it in an individual filter file.
7008 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7009 that creates RFC 2047 encoded "words" labels them as that charset instead
7010 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7011 expansion operator; (ii) when Exim creates a From: line for a local
7012 message; (iii) when a header line is rewritten to include a "phrase" part.
7014 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7015 buggy, causing it to skip the first lines of messages whose message ID
7016 ended in 'D'. This would not have bitten before Exim release 4.14, because
7017 message IDs were unlikely to end in 'D' before then. The effect was to have
7018 incorrect size information for certain domains.
7020 45. #include "config.h" was missing at the start of the crypt16.c module. This
7021 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7024 46. If there was a timeout during a "random" callout check, Exim treated it as
7025 a failure of the random address, and carried on sending RSET and the real
7026 address. If the delay was just some slowness somewhere, the response to the
7027 original RCPT would be taken as a response to RSET and so on, causing
7028 mayhem of various kinds.
7030 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7031 when I implemented it. It didn't allow for the fact that some option values
7032 may legitimately be negative (e.g. size_addition), and it didn't even do
7033 the right test for positive values.
7035 48. Domain names in DNS records are case-independent. Exim always looks them up
7036 in lower case. Some resolvers return domain names in exactly the case they
7037 appear in the zone file, that is, they may contain uppercase letters. Not
7038 all resolvers do this - some return always lower case. Exim was treating a
7039 change of case by a resolver as a change of domain, similar to a widening
7040 of a domain abbreviation. This triggered its re-routing code and so it was
7041 trying to route what was effectively the same domain again. This normally
7042 caused routing to fail (because the router wouldn't handle the domain
7043 twice). Now Exim checks for this case specially, and just changes the
7044 casing of the domain that it ultimately uses when it transmits the message
7047 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7050 50. If a filter generated a file delivery with a non-absolute name (possible if
7051 no home directory exists for the router), the forbid_file option was not
7054 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7055 the existing equality matching.
7057 52. Exim was using ints instead of ino_t variables in some places where it was
7058 dealing with inode numbers.
7060 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7061 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7062 called TMPDIR, and if it finds it is different, it changes its value.
7064 54. The smtp_printf() function is now made available to local_scan() so
7065 additional output lines can be written before returning. There is also an
7066 smtp_fflush() function to enable the detection of a dropped connection.
7067 The variables smtp_input and smtp_batched_input are exported to
7070 55. Changed the default runtime configuration: the message "Unknown user"
7071 has been removed from the ACL, and instead placed on the localuser router,
7072 using the cannot_route_message feature. This means that any verification
7073 failures that generate their own messages won't get overridden. Similarly,
7074 the "Unrouteable address" message that was in the ACL for unverifiable
7075 relay addresses has also been removed.
7077 56. Added hosts_avoid_esmtp to the smtp transport.
7079 57. The exicyclog script was not checking for the esoteric option
7080 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7081 will work only if exicyclog is run under the appropriate euid.
7083 58. Following a discussion on the list, the rules by which Exim recognises line
7084 endings on incoming messages have been changed. The -dropcr and drop_cr
7085 options are now no-ops, retained only for backwards compatibility. The
7086 following line terminators are recognized: LF CRLF CR. However, special
7087 processing applies to CR:
7089 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7090 nor a local message in the state where . is a terminator.
7092 (ii) If a bare CR is encountered in a header line, an extra space is added
7093 after the line terminator so as not to end the header. The reasoning
7094 behind this is that bare CRs in header lines are most likely either
7095 to be mistakes, or people trying to play silly games.
7097 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7098 was being incorrectly given as 18 bytes larger than it should have been.
7099 This is a VOB (very old bug).
7101 60. This may never have affected anything current, but just in case it has:
7102 When the local host is found other than at the start of a list of hosts,
7103 the local host, those with the same MX, and any that follow, are discarded.
7104 When the list in question was part of a longer list of hosts, the following
7105 hosts (not currently being processed) were also being discarded. This no
7106 longer happens. I'm not sure if this situation could ever has previously
7109 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7112 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7113 header. This is something that is recommended in a new Internet Draft, and
7114 is something that is documented as being done by Sendmail. There are two
7115 possible values. For messages generated by the autoreply transport, Exim
7118 Auto-Submitted: auto-replied
7120 whereas for all other generated messages (e.g. bounces) it adds
7122 Auto-Submitted: auto-generated
7124 63. The "personal" condition in filters now includes a test for the
7125 Auto-Submitted: header. If it contains the string "auto-" the message it
7126 not considered personal.
7128 64. Added rcpt_include_affixes as a generic transport option.
7130 65. Added queue_only_override (default true).
7132 66. Added the syslog_duplication option.
7134 67. If what should have been the first header line of a message consisted of
7135 a space followed by a colon, Exim was mis-interpreting it as a header line.
7136 It isn't of course - it is syntactically invalid and should therefore be
7137 treated as the start of the message body. The misbehaviour could have
7138 caused a number of strange effects, including loss of data in subsequent
7139 header lines, and spool format errors.
7141 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7142 client host had authenticated. This control can now be exercised by an ACL
7143 for more flexibility.
7145 69. By default, callouts do not happen when testing with -bh. There is now a
7146 variant, -bhc, which does actually run the callout code, including
7147 consulting and updating the callout cache.
7149 70. Added support for saslauthd authentication, courtesy of Alexander
7152 71. If statvfs() failed on the spool or log directories while checking their
7153 size for availability, Exim confusingly gave the error "space shortage".
7154 Furthermore, in debugging mode it crashed with a floating point exception.
7155 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7156 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7157 serious problem, Exim now writes to the main and panic logs when this
7158 happens, with details of the failure. It then refuses to accept the
7159 incoming message, giving the message "spool directory problem" or "log
7160 directory problem" with a 421 code for SMTP messages.
7162 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7163 0, 1, and 2 exist, because some OS complain for execs without them (see
7164 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7165 descriptors. However, the code omitted to check that the open succeeded,
7166 causing mysterious errors if for some reason the permissions on /dev/null
7167 got screwed. Now Exim writes a message to the main and panic logs, and
7168 bombs out if it can't open /dev/null.
7170 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7171 interact so that it is all more flexible. It is supposed to remain
7172 backwards compatible. Also added extra_local_interfaces.
7174 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7175 to bomb out with an assertion failure - to the client this appears as a
7176 connection drop. This problem occurs in the part of the code that was taken
7177 from the Samba project. Fortunately, the assertion is in a very simple
7178 function, so I have fixed this by reproducing the function inline in the
7179 one place where it is called, and arranging for authentication to fail
7180 instead of killing the process with assert().
7182 75. The SPA client code was not working when the server requested OEM rather
7183 than Unicode encoding.
7185 76. Added code to make require_files with a specific uid setting more usable in
7186 the case where statting the file as root fails - usually a non-root-mounted
7187 NFS file system. When this happens and the failure is EACCES, Exim now
7188 forks a subprocess and does the per-uid checking as the relevant uid.
7190 77. Added process_log_path.
7192 78. If log_file_path was not explicitly set, a setting of check_log_space or
7193 check_log_inodes was ignored.
7195 79. If a space check for the spool or log partitions fails, the incident is now
7196 logged. Of course, in the latter case the data may get lost...
7198 80. Added the %p formatting code to string_format() so that it can be used to
7199 print addresses in debug_print(). Adjusted all the address printing in the
7200 debugging in store.c to use %p rather than %d.
7202 81. There was a concern that a line of code in smtp_in.c could overflow a
7203 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7204 initially expressed, the concern was not well-founded, because trailing
7205 spaces are removed early. However, if the trailing spaces were followed by
7206 a NULL, they did not get removed, so the overflow was possible. Two fixes
7209 (a) I re-wrote the offending code in a cleaner fashion.
7210 (b) If an incoming SMTP command contains a NULL character, it is rejected
7213 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7214 runs initgroups(), so that if the Exim user is in any additional groups,
7215 they will be used during message reception.
7221 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob