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.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
68 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
69 When redis_servers is set to a list of > 1 element, and the Redis servers
70 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
71 case of MOVED into a DEFER case instead, thus moving the query onto the
72 next server in the list. For a cluster of N elements, all N servers must
73 be defined in redis_servers.
75 GF/02 Catch and remove uninitialized value warning in exiqsumm
76 Check for existence of @ARGV before looking at $ARGV[0]
78 JH/01 Replace the store_release() internal interface with store_newblock(),
79 which internalises the check required to safely use the old one, plus
80 the allocate and data copy operations duplicated in both (!) of the
83 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
84 modifier. This matches the restriction on the commandline.
86 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
87 Previously only the last row was returned.
89 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
90 we assumed that tags in the header were well-formed, and parsed the
91 element content after inspecting only the first char of the tag.
92 Assumptions at that stage could crash the receive process on malformed
95 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
96 While running the DKIM ACL we operate on the Permanent memory pool so that
97 variables created with "set" persist to the DATA ACL. Also (at any time)
98 DNS lookups that fail create cache records using the Permanent pool. But
99 expansions release any allocations made on the current pool - so a dnsdb
100 lookup expansion done in the DKIM ACL releases the memory used for the
101 DNS negative-cache, and bad things result. Solution is to switch to the
102 Main pool for expansions.
103 While we're in that code, add checks on the DNS cache during store_reset,
104 active in the testsuite.
105 Problem spotted, and debugging aided, by Wolfgang Breyha.
107 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
108 When none of the hosts presented to a transport match an already-open
109 connection, close it and proceed with the list. Previously we would
110 queue the message. Spotted by Lena with Yahoo, probably involving
113 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
114 Previously a spurious "250 OK id=" response was appended to the proper
117 JH/08 The "support for" informational output now, which built with Content
118 Scanning support, has a line for the malware scanner interfaces compiled
119 in. Interface can be individually included or not at build time.
121 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
122 by the template makefile "src/EDITME". The "STREAM" support for an older
123 ClamAV interface method is removed.
125 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
126 rows affected is given instead).
128 JH/11 The runtime Berkeley DB library version is now additionally output by
129 "exim -d -bV". Previously only the compile-time version was shown.
131 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
132 SMTP connection. Previously, when one had more receipients than the
133 first, an abortive onward connection was made. Move to full support for
134 multiple onward connections in sequence, handling cutthrough connection
135 for all multi-message initiating connections.
137 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
138 routers. Previously, a multi-recipient message would fail to match the
139 onward-connection opened for the first recipient, and cause its closure.
141 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
142 a timeout on read on a GnuTLS initiating connection, resulting in the
143 initiating connection being dropped. This mattered most when the callout
144 was marked defer_ok. Fix to keep the two timeout-detection methods
147 JH/15 Relax results from ACL control request to enable cutthrough, in
148 unsupported situations, from error to silently (except under debug)
149 ignoring. This covers use with PRDR, frozen messages, queue-only and
152 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
154 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
155 metadata, resulting in a crash in free().
157 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
158 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
159 Broken also in d185889f4, with init system revamp.
161 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
162 Previously we abruptly closed the connection after reading a malware-
163 found indication; now we go on to read the "scan ok" response line,
166 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
167 ACL. Previously, a crash would result.
169 JH/19 Speed up macro lookups during configuration file read, by skipping non-
170 macro text after a replacement (previously it was only once per line) and
171 by skipping builtin macros when searching for an uppercase lead character.
173 JH/20 DANE support moved from Experimental to mainline. The Makefile control
174 for the build is renamed.
176 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
177 was allocated for every new TLS startup, meaning one per message. Fix
178 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
180 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
181 reported the original. Fix to report (as far as possible) the ACL
182 result replacing the original.
184 JH/23 Fix memory leak during multi-message connections using STARTTLS under
185 OpenSSL. Certificate information is loaded for every new TLS startup,
186 and the resources needed to be freed.
188 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
190 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
193 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
194 DATA response info to the (existing) per-recipient response info for
195 the "C=" log element. It can have useful tracking info from the
196 destination system. Patch from Simon Arlott.
198 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
199 length value. Previously this would segfault.
201 HS/02 Support Avast multiline protoocol, this allows passing flags to
202 newer versions of the scanner.
204 JH/28 Ensure that variables possibly set during message acceptance are marked
205 dead before release of memory in the daemon loop. This stops complaints
206 about them when the debug_store option is enabled. Discovered specifically
207 for sender_rate_period, but applies to a whole set of variables.
208 Do the same for the queue-runner and queue-list loops, for variables set
209 from spool message files. Do the same for the SMTP per-message loop, for
210 certain variables indirectly set in ACL operations.
212 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
213 as a multi-recipient message from a mailinglist manager). The coding had
214 an arbitrary cutoff number of characters while checking for more input;
215 enforced by writing a NUL into the buffer. This corrupted long / fast
216 input. The problem was exposed more widely when more pipelineing of SMTP
217 responses was introduced, and one Exim system was feeding another.
218 The symptom is log complaints of SMTP syntax error (NUL chars) on the
219 receiving system, and refused recipients seen by the sending system
220 (propating to people being dropped from mailing lists).
221 Discovered and pinpointed by David Carter.
223 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
224 replaced by the ${authresults } expansion.
226 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
228 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
229 allows proper process termination in container environments.
231 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
232 Previously the "final dot" had a newline after it; ensure it is CR,LF.
234 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
235 and "err_perm", deprecated since 4.83 when the RFC-defined words
236 "temperror" and "permerror" were introduced.
238 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
239 transport-filters or DKIM-signing. The restriction was lost in the
240 consolidation of verify-callout and delivery SMTP handling.
241 Extend the restriction to also cover ARC-signing.
243 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
244 in defer=pass mode supply a 450 to the initiator. Previously the message
247 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
248 tls_require_ciphers is used as before.
250 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
251 "pass_unscanned". Only tmpfails from the scanner are written to
252 the paniclog, as they may require admin intervention (permission
253 denied, license issues). Other scanner errors (like decompression
254 bombs) do not cause a paniclog entry.
256 JH/36 Fix reinitialisation of DKIM logging variable between messages.
257 Previously it was possible to log spurious information in receive log
260 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
261 triggered odd behaviour from Outlook Express clients.
263 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
266 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
267 since the IETF WG has not yet settled on that versus the original
268 "bare" representation.
270 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
271 Previously the millisecond value corrupted the output.
272 Fix also for syslog_pid=no and log_selector +pid, for which the pid
273 corrupted the output.
279 JH/01 Rework error string handling in TLS interface so that the caller in
280 more cases is responsible for logging. This permits library-sourced
281 string to be attached to addresses during delivery, and collapses
282 pairs of long lines into single ones.
284 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
285 during configuration. Wildcards are allowed and expanded.
287 JH/02 Rework error string handling in DKIM to pass more info back to callers.
288 This permits better logging.
290 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
291 do not close it down and have the child transport start it up again on
292 the passed-on TCP connection. Instead, proxy the child (and any
293 subsequent ones) for TLS via a unix-domain socket channel. Logging is
294 affected: the continued delivery log lines do not have any DNSSEC, TLS
295 Certificate or OCSP information. TLS cipher information is still logged.
297 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
298 identical IP addresses on different listening ports. Will also affect
301 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
302 add noisy ifdef guards to special-case this sillyness.
303 Patch from Bernd Kuhls.
305 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
306 than 255 are no longer allowed.
308 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
309 Disable the session-cache too, which might reduce our load. Since we
310 currrectly use a new context for every connection, both as server and
311 client, there is no benefit for these.
312 GnuTLS appears to not support tickets server-side by default (we don't
313 call gnutls_session_ticket_enable_server()) but client side is enabled
314 by default on recent versions (3.1.3 +) unless the PFS priority string
317 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
318 <https://reproducible-builds.org/specs/source-date-epoch/>.
320 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
321 the check for any unsuccessful recipients did not notice the limit, and
322 erroneously found still-pending ones.
324 JH/08 Pipeline CHUNKING command and data together, on kernels that support
325 MSG_MORE. Only in-clear (not on TLS connections).
327 JH/09 Avoid using a temporary file during transport using dkim. Unless a
328 transport-filter is involved we can buffer the headers in memory for
329 creating the signature, and read the spool data file once for the
330 signature and again for transmission.
332 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
333 4.77 as the kernel support then wasn't solid, having issues in 64bit
334 mode. Now, it's been long enough. Add support for FreeBSD also.
336 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
337 case where the routing stage had gathered several addresses to send to
338 a host before calling the transport for the first, we previously failed
339 to close down TLS in the old transport process before passing the TCP
340 connection to the new process. The new one sent a STARTTLS command
341 which naturally failed, giving a failed delivery and bloating the retry
342 database. Investigation and fix prototype from Wolfgang Breyha.
344 JH/12 Fix check on SMTP command input synchronisation. Previously there were
345 false-negatives in the check that the sender had not preempted a response
346 or prompt from Exim (running as a server), due to that code's lack of
347 awareness of the SMTP input buffering.
349 PP/04 Add commandline_checks_require_admin option.
350 Exim drops privileges sanely, various checks such as -be aren't a
351 security problem, as long as you trust local users with access to their
352 own account. When invoked by services which pass untrusted data to
353 Exim, this might be an issue. Set this option in main configuration
354 AND make fixes to the calling application, such as using `--` to stop
357 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
358 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
359 responses to those, into a single TLS record each way (this usually means
360 a single packet). As a side issue, smtp_enforce_sync now works on TLS
363 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
364 affects you only if you're dancing at the edge of the param size limits.
365 If you are, and this message makes sense to you, then: raise the
366 configured limit or use OpenSSL 1.1. Nothing we can do for older
369 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
370 cmdline element to get the documented default one. Previously it was
373 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
374 are unsuable for non-message-specific reasons - eg. connection timeout,
375 banner-time rejection.
377 JH/16 Fix logging of delivery remote port, when specified by router, under
380 PP/06 Repair manualroute's ability to take options in any order, even if one
381 is the name of a transport.
384 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
386 JH/17 Change the list-building routines interface to use the expanding-string
387 triplet model, for better allocation and copying behaviour.
389 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
390 Previously it was constructed the first time a possibly-matching string
391 was met in the configuration file input during startup; now it is done
394 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
395 compatible one, to avoid the (poorly documented) possibility of a config
396 file in the working directory redirecting the DB files, possibly correpting
397 some existing file. CVE-2017-10140 assigned for BDB.
399 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
400 cache-hot. Previously, although the result was properly cached, the
401 initial verify call returned a defer.
403 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
404 the main verify for receipient in uncached-mode.
406 JH/22 Retire historical build files to an "unsupported" subdir. These are
407 defined as "ones for which we have no current evidence of testing".
409 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
410 if present. Previously it was ignored.
412 JH/24 Start using specified-initialisers in C structure init coding. This is
413 a C99 feature (it's 2017, so now considered safe).
415 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
416 if was a fixed-sized field and bitmask ops via macros; it is now more
419 PP/07 GitHub PR 56: Apply MariaDB build fix.
420 Patch provided by Jaroslav Škarvada.
422 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
423 during Coverity cleanups [4.87 JH/47]
424 Diagnosis and fix provided by Michael Fischer v. Mollard.
426 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
427 the right size to place the terminating semicolon on its own folded
428 line, the header hash was calculated to an incorrect value thanks to
429 the (relaxed) space the fold became.
431 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
432 and confused the parent.
434 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
435 which could crash as a result. This could lead to undeliverable messages.
437 JH/28 Logging: "next input sent too soon" now shows where input was truncated
440 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
441 matters on fast-turnover and PID-randomising systems, which were getting
442 out-of-order delivery.
444 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
445 a possibly-overlapping copy. The symptom was that "Remote host closed
446 connection in response to HELO" was logged instead of the actual 4xx
449 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
450 Previously only that bufferd was discarded, resulting in SYMTP command
453 JH/32 DKIM: when a message has multiple signatures matching an identity given
454 in dkim_verify_signers, run the dkim acl once for each. Previously only
455 one run was done. Bug 2189.
457 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
458 "panic the current process" to "deliberately defer". The panic log is
459 still written with the problem list name; the mail and reject logs now
460 get a temp-reject line for the message that was being handled, saying
461 something like "domains check lookup or other defer". The SMTP 451
462 message is still "Temporary local problem".
464 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
465 A crafted sequence of BDAT commands could result in in-use memory beeing
466 freed. CVE-2017-16943.
468 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
469 from SMTP input. Previously it was always done; now only done for DATA
470 and not BDAT commands. CVE-2017-16944.
472 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
473 to the message (such as an overlong header line). Previously this was
474 not done and we did not exit BDAT mode. Followon from the previous item
475 though a different problem.
481 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
482 than -2003 did; needs libidn2 in addition to libidn.
484 JH/02 The path option on a pipe transport is now expanded before use.
486 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
487 Patch provided by "Björn", documentation fix added too.
489 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
490 missing a wire-to-host endian conversion.
492 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
493 close after a BDAT command line could be taken as a following command,
494 giving a synch failure. Fix by only checking for synch immediately
495 before acknowledging the chunk.
497 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
498 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
499 macro. Patches provided by Josh Soref.
501 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
502 Previously we did not; the RFC seems ambiguous and VRFY is not listed
503 by IANA as a service extension. However, John Klensin suggests that we
506 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
507 the dkim code may be unix-mode line endings rather than smtp wire-format
508 CRLF, so prepend a CR to any bare LF.
510 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
511 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
513 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
514 the first were themselves being wrongly included in the feed into dkim
515 processing; with most chunk sizes in use this resulted in an incorrect
516 body hash calculated value.
518 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
519 DKIM signature block, for verification. Although advised against by
520 standards it is specifically not ruled illegal.
522 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
524 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
525 missing a body hash (the bh= tag).
527 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
528 It seems that HAProxy sends the Proxy Protocol information in clear and
529 only then does a TLS startup, so do the same.
531 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
532 TCP connections (such as for Spamd) unless the daemon successfully set
533 Fast Open mode on its listening sockets. This fixes breakage seen on
534 too-old kernels or those not configured for Fast Open, at the cost of
535 requiring both directions being enabled for TFO, and TFO never being used
536 by non-daemon-related Exim processes.
538 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
539 endings, at least on the first header line. Try to canonify any that get
540 past that check, despite the cost.
542 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
543 now limited to an arbitrary five deep, while parsing addresses with the
544 strip_excess_angle_brackets option enabled.
546 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
547 instead leave the unprompted TLS handshake in socket buffer for the
548 TLS library to consume.
550 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
552 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
554 JH/16 Drop variables when they go out of scope. Memory management drops a whole
555 region in one operation, for speed, and this leaves assigned pointers
556 dangling. Add checks run only under the testsuite which checks all
557 variables at a store-reset and panics on a dangling pointer; add code
558 explicitly nulling out all the variables discovered. Fixes one known
559 bug: a transport crash, where a dangling pointer for $sending_ip_address
560 originally assigned in a verify callout, is re-used.
562 PP/06 Drop '.' from @INC in various Perl scripts.
564 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
566 PP/08 Reduce a number of compilation warnings under clang; building with
567 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
568 should be warning-free.
570 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
572 HS/01 Fix portability problems introduced by PP/08 for platforms where
573 realloc(NULL) is not equivalent to malloc() [SunOS et al].
575 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
576 chunk. This allows us to accept broken chunked messages. We need a more
577 general solution here.
579 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
580 already-broken messages in the queue.
582 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
584 JH/19 Fix reference counting bug in routing-generated-address tracking.
590 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
591 supports it and a size is available (ie. the sending peer gave us one).
593 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
594 years of being deprecated). The replacements are the ACLs
595 acl_smtp_mime and acl_not_smtp_mime.
597 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
598 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
599 or even an in-clear connection were permitted. Now, if the host lookup
600 was dnssec and dane was requested then the host is only used if the
601 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
602 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
603 if one fails this test.
604 This means that a poorly-configured remote DNS will make it incommunicado;
605 but it protects against a DNS-interception attack on it.
607 JH/04 Bug 1810: make continued-use of an open smtp transport connection
608 non-noisy when a race steals the message being considered.
610 JH/05 If main configuration option tls_certificate is unset, generate a
611 self-signed certificate for inbound TLS connections.
613 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
614 in rewrites and routers.
616 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
617 and logged a warning sing 4.83; now they are a configuration file error.
619 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
620 (lacking @domain). Apply the same qualification processing as RCPT.
622 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
624 JH/10 Support ${sha256:} applied to a string (as well as the previous
627 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
628 a cutthrough deliver is pending, as we always want to make a connection.
629 This also avoids re-routing the message when later placing the cutthrough
630 connection after a verify cache hit.
631 Do not update it with the verify result either.
633 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
634 when routing results in more than one destination address.
636 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
637 signing (which inhibits the cutthrough capability). Previously only
638 the presence of an option was tested; now an expansion evaluating as
639 empty is permissible (obviously it should depend only on data available
640 when the cutthrough connection is made).
642 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
643 the relevant preceding SMTP command did not note the pipelining mode.
645 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
646 Previously they were not counted.
648 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
649 as one having no matching records. Previously we deferred the message
650 that needed the lookup.
652 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
653 distinguished as "(=".
655 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
656 for missing MX records. Previously it only worked for missing A records.
658 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
660 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
661 after the data-go-ahead and data-ack. Patch from Jason Betts.
663 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
664 even for a "none" policy. Patch from Tony Meyer.
666 JH/22 Fix continued use of a connection for further deliveries. If a port was
667 specified by a router, it must also match for the delivery to be
670 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
671 When one of the recipients of a message was unsuitable for the connection
672 (has no matching addresses), we lost track of needing to mark it
673 deferred. As a result mail would be lost.
675 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
677 JH/25 Decoding ACL controls is now done using a binary search; the source code
678 takes up less space and should be simpler to maintain. Merge the ACL
679 condition decode tables also, with similar effect.
681 JH/26 Fix problem with one_time used on a redirect router which returned the
682 parent address unchanged. A retry would see the parent address marked as
683 delivered, so not attempt the (identical) child. As a result mail would
686 JH/27 Fix a possible security hole, wherein a process operating with the Exim
687 UID can gain a root shell. Credit to http://www.halfdog.net/ for
688 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
691 JH/28 Enable {spool,log} filesystem space and inode checks as default.
692 Main config options check_{log,spool}_{inodes,space} are now
693 100 inodes, 10MB unless set otherwise in the configuration.
695 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
696 Previously it only applied to the main-section connection policy
699 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
701 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
702 by me. Added RFC7919 DH primes as an alternative.
704 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
705 are not in the system include path.
707 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
708 GnuTLS, when a session startup failed (eg because the client disconnected)
709 Exim did stdio operations after fclose. This was exposed by a recent
710 change which nulled out the file handle after the fclose.
712 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
713 signed directly by the cert-signing cert, rather than an intermediate
714 OCSP-signing cert. This is the model used by LetsEncrypt.
716 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
718 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
719 an incoming connection.
721 HS/02 Bug 1802: Do not half-close the connection after sending a request
724 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
725 fallback to "prime256v1".
727 JH/34 SECURITY: Use proper copy of DATA command in error message.
728 Could leak key material. Remotely exploitable. CVE-2016-9963.
734 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
735 and 3.4.4 - once the server is enabled to respond to an OCSP request
736 it does even when not requested, resulting in a stapling non-aware
737 client dropping the TLS connection.
739 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
740 support variable-length bit vectors. No functional change.
742 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
743 The I= interface field on outgoing lines is now after the H= remote
744 host field, same as incoming lines. There is a separate
745 outgoing_interface log selector which allows you to disable the
748 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
749 If not running log_selector +smtp_connection the mainlog would be held
750 open indefinitely after a "too many connections" event, including to a
751 deleted file after a log rotate. Leave the per net connection logging
752 leaving it open for efficiency as that will be quickly detected by the
753 check on the next write.
755 HS/01 Bug 1671: Fix post transport crash.
756 Processing the wait-<transport> messages could crash the delivery
757 process if the message IDs didn't exist for some reason. When
758 using 'split_spool_directory=yes' the construction of the spool
759 file name failed already, exposing the same netto behaviour.
761 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
762 mime_regex ACL conditions.
764 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
765 to DSN fail messages (bounces): remote IP, remote greeting, remote response
766 to HELO, local diagnostic string.
768 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
769 log line to debug. Even when configured with a tls authenticator many
770 client connections are expected to not authenticate in this way, so
771 an authenticate fail is not an error.
773 HS/02 Add the Exim version string to the process info. This way exiwhat
774 gives some more detail about the running daemon.
776 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
777 matter for fast-change records such as DNSBLs.
779 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
780 retry record, even if constant. There may be multiple transports with
781 different interface settings and the retry behaviour needs to be kept
784 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
786 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
788 JH/10 Bug 840: fix log_defer_output option of pipe transport
790 JH/11 Bug 830: use same host for all RCPTS of a message, even under
791 hosts_randomize. This matters a lot when combined with mua_wrapper.
793 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
794 ${quote_pgsql:<string>} operator.
796 JH/13 Bug 1708: avoid misaligned access in cached lookup.
798 JH/14 Change header file name for freeradius-client. Relevant if compiling
799 with Radius support; from the Gentoo tree and checked under Fedora.
801 JH/15 Bug 1712: Introduce $prdr_requested flag variable
803 JH/16 Bug 1714: Permit an empty string as expansion result for transport
804 option transport_filter, meaning no filtering.
806 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
808 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
809 defaults to "*" (all hosts). The variable is now available when not built
810 with TLS, default unset, mainly to enable keeping the testsuite sane.
811 If a server certificate is not supplied (via tls_certificate) an error is
812 logged, and clients will find TLS connections fail on startup. Presumably
813 they will retry in-clear.
814 Packagers of Exim are strongly encouraged to create a server certificate
815 at installation time.
817 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
818 with the $config_file variable.
820 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
821 in transport context, after the attempt, and per-recipient. The latter type
822 is per host attempted. The event data is the error message, and the errno
823 information encodes the lookup type (A vs. MX) used for the (first) host,
824 and the trailing two digits of the smtp 4xx response.
826 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
827 to write to mainlog (or rejectlog, paniclog) in the window between file
828 creation and permissions/ownership being changed. Particularly affects
829 installations where exicyclog is run as root, rather than exim user;
830 result is that the running daemon panics and dies.
832 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
834 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
835 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
836 "pri" and "weight". Note that the previous implicit priority given by the
837 list order is no longer honoured.
839 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
842 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
843 by defining SUPPORT_SOCKS.
845 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
846 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
847 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
848 variables are renamed to proxy_{local,external}_{address,port}.
850 JH/27 Move Internationalisation support from Experimental to mainline, enabled
851 for a build by defining SUPPORT_I18N
853 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
854 of the query string, and make ${quote_redis:} do that quoting.
856 JH/29 Move Events support from Experimental to mainline, enabled by default
857 and removable for a build by defining DISABLE_EVENT.
859 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
861 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
862 cached by the daemon.
864 JH/32 Move Redis support from Experimental to mainline, enabled for a build
865 by defining LOOKUP_REDIS. The libhiredis library is required.
867 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
868 keys are given for lookup.
870 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
871 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
872 only supported when built with TLS support. The PolarSSL SHA routines
873 are still used when the TLS library is too old for convenient support.
875 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
876 openssl_options), for security. OpenSSL forces this from version 1.1.0
877 server-side so match that on older versions.
879 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
880 allocation for $value could be released as the expansion processing
881 concluded, but leaving the global pointer active for it.
883 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
884 and to use the domains and local_parts ACL conditions.
886 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
887 incorrectly not doubled on cutthrough transmission, hence seen as a
888 body-termination at the receiving system - resulting in truncated mails.
889 Commonly the sender saw a TCP-level error, and retransmitted the message
890 via the normal store-and-forward channel. This could result in duplicates
891 received - but deduplicating mailstores were liable to retain only the
892 initial truncated version.
894 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
896 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
898 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
899 we're in there, support oversigning also; bug 1309.
901 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
903 HS/04 Add support for keep_environment and add_environment options.
905 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
906 either intentional arithmetic overflow during PRNG, or testing config-
909 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
910 delivery resulted in actual delivery. Cancel cutthrough before DATA
913 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
914 rejects a recipient: pass the reject to the originator.
916 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
917 Many were false-positives and ignorable, but it's worth fixing the
920 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
921 for the new environment-manipulation done at startup. Move the routines
922 from being local to tls.c to being global via the os.c file.
924 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
925 an extract embedded as result-arg for a map, the first arg for extract
926 is unavailable so we cannot tell if this is a numbered or keyed
927 extraction. Accept either.
933 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
936 JH/02 The smtp transport option "multi_domain" is now expanded.
938 JH/03 The smtp transport now requests PRDR by default, if the server offers
941 JH/04 Certificate name checking on server certificates, when exim is a client,
942 is now done by default. The transport option tls_verify_cert_hostnames
943 can be used to disable this per-host. The build option
944 EXPERIMENTAL_CERTNAMES is withdrawn.
946 JH/05 The value of the tls_verify_certificates smtp transport and main options
947 default to the word "system" to access the system default CA bundle.
948 For GnuTLS, only version 3.0.20 or later.
950 JH/06 Verification of the server certificate for a TLS connection is now tried
951 (but not required) by default. The verification status is now logged by
952 default, for both outbound TLS and client-certificate supplying inbound
955 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
958 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
959 Status Notification (bounce) messages are now MIME format per RFC 3464.
960 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
961 under the control of the dsn_advertise_hosts option, and routers may
962 have a dsn_lasthop option.
964 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
965 default, modifiable by a malware= option. The list separator for
966 the options can now be changed in the usual way. Bug 68.
968 JH/10 The smtp_receive_timeout main option is now expanded before use.
970 JH/11 The incoming_interface log option now also enables logging of the
971 local interface on delivery outgoing connections.
973 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
974 if the interface and destination host and port all match.
976 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
979 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
980 Patch from Andrew Lewis.
982 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
983 now supports optional time-restrictions, weighting, and priority
984 modifiers per server. Patch originally by <rommer@active.by>.
986 JH/16 The spamd_address main option now supports a mixed list of local
987 and remote servers. Remote servers can be IPv6 addresses, and
988 specify a port-range.
990 JH/17 Bug 68: The spamd_address main option now supports an optional
991 timeout value per server.
993 JH/18 Bug 1581: Router and transport options headers_add/remove can
994 now have the list separator specified.
996 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
999 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1002 JH/21 Support for the A6 type of dns record is withdrawn.
1004 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1005 rather than the verbs used.
1007 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1008 from 255 to 1024 chars.
1010 JH/24 Verification callouts now attempt to use TLS by default.
1012 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1013 are generic router options now. The defaults didn't change.
1015 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1016 Original patch from Alexander Shikoff, worked over by JH.
1018 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1019 files if ZCAT_COMMAND is not executable.
1021 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1023 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1025 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1026 Normally benign, it bites when the pair was led to by a CNAME;
1027 modern usage is to not canonicalize the domain to a CNAME target
1028 (and we were inconsistent anyway for A-only vs AAAA+A).
1030 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1032 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1033 when evaluating $sender_host_dnssec.
1035 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1036 $sender_helo_dnssec variable.
1038 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1040 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1042 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1044 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1045 documented as working, but never had. Support all but $spam_report.
1047 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1048 added for tls authenticator.
1050 HS/03 Add perl_taintmode main config option
1056 TL/01 When running the test suite, the README says that variables such as
1057 no_msglog_check are global and can be placed anywhere in a specific
1058 test's script, however it was observed that placement needed to be near
1059 the beginning for it to behave that way. Changed the runtest perl
1060 script to read through the entire script once to detect and set these
1061 variables, reset to the beginning of the script, and then run through
1062 the script parsing/test process like normal.
1064 TL/02 The BSD's have an arc4random API. One of the functions to induce
1065 adding randomness was arc4random_stir(), but it has been removed in
1066 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1067 function when detected.
1069 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1070 cause callback expansion.
1072 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1073 syntax errors in an expansion can be treated as a string instead of
1074 logging or causing an error, due to the internal use of bool_lax
1075 instead of bool when processing it.
1077 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1078 server certificates when making smtp deliveries.
1080 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1082 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1084 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1086 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1087 Merged patch from Sebastian Wiedenroth.
1089 JH/05 Fix results-pipe from transport process. Several recipients, combined
1090 with certificate use, exposed issues where response data items split
1091 over buffer boundaries were not parsed properly. This eventually
1092 resulted in duplicates being sent. This issue only became common enough
1093 to notice due to the introduction of connection certificate information,
1094 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1096 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1097 size buffer was used, resulting in syntax errors when an expansion
1100 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1101 version 3.3.6 or later.
1103 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1104 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1105 both become "event_action", the variables become $event_name, $event_data
1106 and $event_defer_errno. There is a new variable $verify_mode, usable in
1107 routers, transports and related events. The tls:cert event is now also
1108 raised for inbound connections, if the main configuration event_action
1111 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1112 early OCSP support, but no stapling (appears to be less than 1.0.0).
1114 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1115 server certificate names available under the smtp transport option
1116 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1119 JH/10 Time-related extraction expansions from certificates now use the main
1120 option "timezone" setting for output formatting, and are consistent
1121 between OpenSSL and GnuTLS compilations. Bug 1541.
1123 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1124 encoded parameter in the incoming message. Bug 1558.
1126 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1127 include certificate info, eximon was claiming there were spoolfile
1130 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1132 JH/14 Log delivery-related information more consistently, using the sequence
1133 "H=<name> [<ip>]" wherever possible.
1135 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1136 are problematic for Debian distribution, omit them from the release
1139 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1141 JH/16 Fix string representation of time values on 64bit time_t architectures.
1144 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1145 output list separator was used.
1150 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1151 checkers that were complaining about end of non-void function with no
1154 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1155 This was a regression introduced in 4.83 by another bugfix.
1157 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1159 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1160 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1166 TF/01 Correctly close the server side of TLS when forking for delivery.
1168 When a message was received over SMTP with TLS, Exim failed to clear up
1169 the incoming connection properly after forking off the child process to
1170 deliver the message. In some situations the subsequent outgoing
1171 delivery connection happened to have the same fd number as the incoming
1172 connection previously had. Exim would try to use TLS and fail, logging
1173 a "Bad file descriptor" error.
1175 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1176 utilities have not been installed.
1178 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1179 temporary space as the ACL may create new global variables.
1181 TL/01 LDAP support uses per connection or global context settings, depending
1182 upon the detected version of the libraries at build time.
1184 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1185 to extract and use the src ip:port in logging and expansions as if it
1186 were a direct connection from the outside internet. PPv2 support was
1187 updated based on HAProxy spec change in May 2014.
1189 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1191 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1192 Properly escape header and check for NULL return.
1194 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1197 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1199 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1200 characters in header names, implemented as a verify condition.
1201 Contributed by Michael Fischer v. Mollard.
1203 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1204 results permerror and temperror. Previous values are deprecated but
1205 still accepted. In a future release, err_perm and err_temp will be
1206 completely removed, which will be a backward incompatibility if the
1207 ACL tests for either of these two old results. Patch contributed by
1208 user bes-internal on the mailing list.
1210 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1212 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1213 selectors, in both main and reject logs.
1215 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1218 JH/07 Add malware type "sock" for talking to simple daemon.
1220 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1222 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1223 routers/transports under cutthrough routing.
1225 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1226 numbers. Touch up "bool" conditional to keep the same definition.
1228 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1230 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1232 JH/12 Expand items in router/transport headers_add or headers_remove lists
1233 individually rather than the list as a whole. Bug 1452.
1235 Required for reasonable handling of multiple headers_ options when
1236 they may be empty; requires that headers_remove items with embedded
1237 colons must have them doubled (or the list-separator changed).
1239 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1240 view the policy declared in the DMARC record. Currently, $dmarc_status
1241 is a combined value of both the record presence and the result of the
1244 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1246 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1247 dnslookup router and the smtp transport (applying to the forward
1250 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1251 of ldap servers used for a specific lookup. Patch provided by Heiko
1254 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1255 New variable $lookup_dnssec_authenticated for observability.
1257 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1258 Patch submitted by Lars Timman.
1260 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1262 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1263 Requires trusted mode and valid format message id, aborts otherwise.
1264 Patch contributed by Heiko Schlichting.
1266 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1267 certextract with support for various fields. Bug 1358.
1269 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1270 is requested by default, modifiable by smtp transport option
1273 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1274 operate on certificate variables to give certificate fingerprints
1275 Also new ${sha256:cert_variable}.
1277 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1279 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1280 Christian Aistleitner.
1282 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1284 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1285 file. Patch from Wolfgang Breyha.
1287 JH/25 Expand the coverage of the delivery $host and $host_address to
1288 client authenticators run in verify callout. Bug 1476.
1290 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1291 align with daemon_smtp_ports. Bug 72.
1293 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1294 support and error reporting did not work properly.
1296 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1297 and is readable. Patch from Andrew Colin Kissa.
1299 TL/14 Enhance documentation of ${run expansion and how it parses the
1300 commandline after expansion, particularly in the case when an
1301 unquoted variable expansion results in an empty value.
1303 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1305 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1306 by renaming away from it. Addresses GH issue 15, reported by
1309 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1311 TL/15 SECURITY: prevent double expansion in math comparison functions
1312 (can expand unsanitized data). Not remotely exploitable.
1319 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1321 PP/02 Make -n do something, by making it not do something.
1322 When combined with -bP, the name of an option is not output.
1324 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1327 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1328 $sender_host_name and config options to manage this, and basic check
1331 PP/05 DSCP support for outbound connections and control modifier for inbound.
1333 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1334 (Only plugin which currently uses this is kerberos4, which nobody should
1335 be using, but we should make it available and other future plugins might
1336 conceivably use it, even though it would break NAT; stuff *should* be
1337 using channel bindings instead).
1339 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1340 name; added for Sendmail compatibility; requires admin caller.
1341 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1342 just ignore it); requires trusted caller.
1343 Also parse but ignore: -Ac -Am -X<logfile>
1346 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1348 TL/02 Add +smtp_confirmation as a default logging option.
1350 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1351 Patch by Magnus Holmgren from 2007-02-20.
1353 TL/04 Bugzilla 1281 - Spec typo.
1354 Bugzilla 1283 - Spec typo.
1355 Bugzilla 1290 - Spec grammar fixes.
1357 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1359 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1361 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1362 mailing list by by Dmitry Isaikin.
1364 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1366 JH/02 Support "G" suffix to numbers in ${if comparisons.
1368 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1370 NM/01 Bugzilla 1197 - Spec typo
1371 Bugzilla 1196 - Spec examples corrections
1373 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1375 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1376 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1379 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1380 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1382 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1383 "acl {{name}{arg}...}", and optional args on acl condition
1386 JH/05 Permit multiple router/transport headers_add/remove lines.
1388 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1390 JH/07 Avoid using a waiting database for a single-message-only transport.
1391 Performance patch from Paul Fisher. Bugzilla 1262.
1393 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1396 JH/09 Add $headers_added variable, with content from use of ACL modifier
1397 add_header (but not yet added to the message). Bugzilla 199.
1399 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1400 Pulled from Bugzilla 817 by Wolfgang Breyha.
1402 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1404 (nb: this is the same fix as in Exim 4.80.1)
1406 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1409 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1412 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1414 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1415 advertise SMTP AUTH mechanism to us, instead of a generic
1416 protocol violation error. Also, make Exim more robust to bad
1417 data from the Dovecot auth socket.
1419 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1421 When a queue runner is handling a message, Exim first routes the
1422 recipient addresses, during which it prunes them based on the retry
1423 hints database. After that it attempts to deliver the message to
1424 any remaining recipients. It then updates the hints database using
1427 So if a recipient address works intermittently, it can get repeatedly
1428 deferred at routing time. The retry hints record remains fresh so the
1429 address never reaches the final cutoff time.
1431 This is a fairly common occurrence when a user is bumping up against
1432 their storage quota. Exim had some logic in its local delivery code
1433 to deal with this. However it did not apply to per-recipient defers
1434 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1436 This change adds a proper retry rule check during routing so that the
1437 final cutoff time is checked against the message's age. We only do
1438 this check if there is an address retry record and there is not a
1439 domain retry record; this implies that previous attempts to handle
1440 the address had the retry_use_local_parts option turned on. We use
1441 this as an approximation for the destination being like a local
1442 delivery, as in LMTP.
1444 I suspect this new check makes the old local delivery cutoff check
1445 redundant, but I have not verified this so I left the code in place.
1447 TF/02 Correct gecos expansion when From: is a prefix of the username.
1449 Test 0254 submits a message to Exim with the header
1453 When I ran the test suite under the user fanf2, Exim expanded
1454 the header to contain my full name, whereas it should have added
1455 a Resent-Sender: header. It erroneously treats any prefix of the
1456 username as equal to the username.
1458 This change corrects that bug.
1460 GF/01 DCC debug and logging tidyup
1461 Error conditions log to paniclog rather than rejectlog.
1462 Debug lines prefixed by "DCC: " to remove any ambiguity.
1464 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1466 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1467 Bug spotted by Jeremy Harris; was flawed since initial commit.
1468 Would have resulted in OCSP responses post-SNI triggering an Exim
1469 NULL dereference and crash.
1471 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1473 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1474 Bug detection, analysis and fix by Samuel Thibault.
1475 Bugzilla 1331, Debian bug #698092.
1477 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1479 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1480 Server implementation by Todd Lyons, client by JH.
1481 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1482 config variable "prdr_enable" controls whether the server
1483 advertises the facility. If the client requests PRDR a new
1484 acl_data_smtp_prdr ACL is called once for each recipient, after
1485 the body content is received and before the acl_smtp_data ACL.
1486 The client is controlled by both of: a hosts_try_prdr option
1487 on the smtp transport, and the server advertisement.
1488 Default client logging of deliveries and rejections involving
1489 PRDR are flagged with the string "PRDR".
1491 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1492 fclose(). Diagnosis by Todd Lyons.
1494 PP/17 Update configure.default to handle IPv6 localhost better.
1495 Patch by Alain Williams (plus minor tweaks).
1498 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1499 This is now consistent with GnuTLS, and is now documented: the
1500 previous undocumented portable approach to treating the option as
1501 unset was to force an expansion failure. That still works, and
1502 an empty string is now equivalent.
1504 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1505 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1506 not performing validation itself.
1508 PP/20 Added force_command boolean option to pipe transport.
1509 Patch from Nick Koston, of cPanel Inc.
1511 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1514 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1516 PP/21 Fix eximon continuous updating with timestamped log-files.
1517 Broken in a format-string cleanup in 4.80, missed when I repaired the
1518 other false fix of the same issue.
1519 Report and fix from Heiko Schlichting.
1522 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1523 Report from Prashanth Katuri.
1525 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1526 It's SecureTransport, so affects any MacOS clients which use the
1527 system-integrated TLS libraries, including email clients.
1529 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1530 using a MIME ACL for non-SMTP local injection.
1531 Report and assistance in diagnosis by Warren Baker.
1533 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1535 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1537 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1538 last id that failed so it may be referenced in subsequent ACL's.
1540 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1543 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1544 ldap library initialization, allowing self-signed CA's to be
1545 used. Also properly sets require_cert option later in code by
1546 using NULL (global ldap config) instead of ldap handle (per
1547 session). Bug diagnosis and testing by alxgomz.
1549 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1550 the src/util/ subdirectory.
1552 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1553 renamed to Transport Post Delivery Action by Jeremy Harris, as
1556 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1557 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1558 redis_servers = needs to be configured which will be used by the redis
1559 lookup. Patch from Warren Baker, of The Packet Hub.
1561 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1563 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1564 hostname or reverse DNS when processing a host list. Used suggestions
1565 from multiple comments on this bug.
1567 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1569 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1570 Missed a few lines, added it to make the runtest require no keyboard
1573 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1574 contains upper case chars. Make router use caseful_local_part.
1576 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1577 support when GnuTLS has been built with p11-kit.
1583 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1585 This, or similar/improved, will also be change PP/11 of 4.82.
1591 PP/01 Handle short writes when writing local log-files.
1592 In practice, only affects FreeBSD (8 onwards).
1593 Bugzilla 1053, with thanks to Dmitry Isaikin.
1595 NM/01 Bugzilla 949 - Documentation tweak
1597 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1600 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1602 PP/02 Implemented gsasl authenticator.
1604 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1606 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1607 `pkg-config foo` for cflags/libs.
1609 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1610 with rest of GSASL and with heimdal_gssapi.
1612 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1613 `pkg-config foo` for cflags/libs for the TLS implementation.
1615 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1616 properties get this fed in as external SSF. A number of robustness
1617 and debugging improvements to the cyrus_sasl authenticator.
1619 PP/08 cyrus_sasl server now expands the server_realm option.
1621 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1622 Patch by Jeremy Harris.
1624 PP/10 Added dbmjz lookup type.
1626 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1628 PP/12 MAIL args handles TAB as well as SP, for better interop with
1629 non-compliant senders.
1630 Analysis and variant patch by Todd Lyons.
1632 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1633 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1634 Patch from Dirk Mueller <dmueller@suse.com>
1636 PP/13 tls_peerdn now print-escaped for spool files.
1637 Observed some $tls_peerdn in wild which contained \n, which resulted
1638 in spool file corruption.
1640 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1641 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1642 or write after TLS renegotiation, which otherwise led to messages
1645 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1646 as a tracking header (ie: a signed header comes before the signature).
1647 Patch from Wolfgang Breyha.
1649 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1650 comma-sep list; embedded commas doubled.
1652 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1654 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1656 Report and patch from Dmitry Banschikov.
1658 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1659 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1660 failures. We appear to now support TLS1.1+ with Exim.
1662 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1663 lets Exim select keys and certificates based upon TLS SNI from client.
1664 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1665 before an outbound SMTP session. New log_selector, +tls_sni.
1667 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1668 NULL dereference. Report and patch from Alun Jones.
1670 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1671 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1672 Not seeing resolver debug output on NetBSD, but suspect this is a
1673 resolver implementation change.
1675 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1676 Left warnings. Added "eximon gdb" invocation mode.
1678 PP/21 Defaulting "accept_8bitmime" to true, not false.
1680 PP/22 Added -bw for inetd wait mode support.
1682 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1683 locate the relevant includes and libraries. Made this the default.
1685 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1686 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1688 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1689 This may cause build issues on older platforms.
1691 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1692 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1693 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1694 Added SNI support via GnuTLS too.
1695 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1697 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1699 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1700 Applied second patch from Janne, implementing suggestion to default
1701 multiple-strings-in-record handling to match SPF spec.
1703 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1705 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1706 read-only, out of scope).
1707 Patch from Wolfgang Breyha, report from Stuart Northfield.
1709 PP/29 Fix three issues highlighted by clang analyser static analysis.
1710 Only crash-plausible issue would require the Cambridge-specific
1711 iplookup router and a misconfiguration.
1712 Report from Marcin Mirosław.
1714 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1716 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1717 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1718 As part of this, removing so much warning spew let me fix some minor
1719 real issues in debug logging.
1721 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1722 assignment on my part. Fixed.
1724 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1725 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1726 Janne Snabb (who went above and beyond: thank you).
1728 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1729 string otherwise requires a connection and a bunch more work and it's
1730 relatively easy to get wrong. Should also expose TLS library linkage
1733 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1734 64-bit ${eval} (JH/03).
1736 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1737 GNU libc to support some of the 64-bit stuff, should not lead to
1738 conflicts. Defined before os.h is pulled in, so if a given platform
1739 needs to override this, it can.
1741 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1742 protection layer was required, which is not implemented.
1743 Bugzilla 1254, patch from Wolfgang Breyha.
1745 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1746 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1747 tls_dhparam take prime identifiers. Also unbreak combination of
1748 OpenSSL+DH_params+TLSSNI.
1750 PP/39 Disable SSLv2 by default in OpenSSL support.
1756 PP/01 Solaris build fix for Oracle's LDAP libraries.
1757 Bugzilla 1109, patch from Stephen Usher.
1759 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1761 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1764 TF/02 Fix a couple more cases where we did not log the error message
1765 when unlink() failed. See also change 4.74-TF/03.
1767 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1768 lock up or crash if it happened to be inside a call to libc when it
1769 got a SIGUSR1 from exiwhat.
1771 The SIGUSR1 handler appends the current process status to the process
1772 log which is later printed by exiwhat. It used to use the general
1773 purpose logging code to do this, but several functions it calls are
1774 not safe for signals.
1776 The new output code in the SIGUSR1 handler is specific to the process
1777 log, and simple enough that it's easy to inspect for signal safety.
1778 Removing some special cases also simplifies the general logging code.
1779 Removing the spurious timestamps from the process log simplifies
1782 TF/04 Improved ratelimit ACL condition.
1784 The /noupdate option has been deprecated in favour of /readonly which
1785 has clearer semantics. The /leaky, /strict, and /readonly update modes
1786 are mutually exclusive. The update mode is no longer included in the
1787 database key; it just determines when the database is updated. (This
1788 means that when you upgrade Exim will forget old rate measurements.)
1790 Exim now checks that the per_* options are used with an update mode that
1791 makes sense for the current ACL. For example, when Exim is processing a
1792 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1793 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1794 must specify per_mail/readonly. If you omit the update mode it defaults to
1795 /leaky where that makes sense (as before) or /readonly where required.
1797 The /noupdate option is now undocumented but still supported for
1798 backwards compatibility. It is equivalent to /readonly except that in
1799 ACLs where /readonly is required you may specify /leaky/noupdate or
1800 /strict/noupdate which are treated the same as /readonly.
1802 A useful new feature is the /count= option. This is a generalization
1803 of the per_byte option, so that you can measure the throughput of other
1804 aggregate values. For example, the per_byte option is now equivalent
1805 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1807 The per_rcpt option has been generalized using the /count= mechanism
1808 (though it's more complicated than the per_byte equivalence). When it is
1809 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1810 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1811 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1812 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1813 Note that using per_rcpt with a non-readonly update mode in more than
1814 one ACL will cause the recipients to be double-counted. (The per_mail
1815 and per_byte options don't have this problem.)
1817 The handling of very low rates has changed slightly. If the computed rate
1818 is less than the event's count (usually one) then this event is the first
1819 after a long gap. In this case the rate is set to the same as this event's
1820 count, so that the first message of a spam run is counted properly.
1822 The major new feature is a mechanism for counting the rate of unique
1823 events. The new per_addr option counts the number of different
1824 recipients that someone has sent messages to in the last time period. It
1825 behaves like per_rcpt if all the recipient addresses are different, but
1826 duplicate recipient addresses do not increase the measured rate. Like
1827 the /count= option this is a general mechanism, so the per_addr option
1828 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1829 example, measure the rate that a client uses different sender addresses
1830 with the options per_mail/unique=$sender_address. There are further
1831 details in the main documentation.
1833 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1835 TF/06 Removed a few PCRE remnants.
1837 TF/07 Automatically extract Exim's version number from tags in the git
1838 repository when doing development or release builds.
1840 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1841 Bugzilla 879. Patch from Paul Fisher.
1843 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1844 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1847 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1849 PP/05 Variable $av_failed, true if the AV scanner deferred.
1850 Bugzilla 1078. Patch from John Horne.
1852 PP/06 Stop make process more reliably on build failure.
1853 Bugzilla 1087. Patch from Heiko Schlittermann.
1855 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1856 Bugzilla 1089. Patch from Heiko Schlittermann.
1858 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1859 Bugzilla 1131. Patch from Holger Weiß.
1861 PP/09 Handle IPv6 addresses with SPF.
1862 Bugzilla 860. Patch from Wolfgang Breyha.
1864 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1866 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1869 PP/11 match_* no longer expand right-hand-side by default.
1870 New compile-time build option, EXPAND_LISTMATCH_RHS.
1871 New expansion conditions, "inlist", "inlisti".
1873 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1875 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1877 PP/14 fix log_write() format string regression from TF/03.
1878 Bugzilla 1152. Patch from Dmitry Isaikin.
1884 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1886 PP/02 Harmonised TLS library version reporting; only show if debugging.
1887 Layout now matches that introduced for other libraries in 4.74 PP/03.
1889 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1891 PP/04 New "dns_use_edns0" global option.
1893 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1896 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1897 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1899 TK/01 Updated PolarSSL code to 0.14.2.
1900 Bugzilla 1097. Patch from Andreas Metzler.
1902 PP/07 Catch divide-by-zero in ${eval:...}.
1903 Fixes bugzilla 1102.
1905 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1908 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1909 format-string attack -- SECURITY: remote arbitrary code execution.
1911 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1912 time unintentionally subject to list matching rules, letting the header
1913 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1914 arbitrary string expansion). This allowed for information disclosure.
1916 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1917 INT_MIN/-1 -- value coerced to INT_MAX.
1923 NM/01 Workaround for PCRE version dependency in version reporting
1926 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1927 This fixes portability to compilers other than gcc, notably
1928 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1930 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1931 makefiles for portability to HP-UX and POSIX correctness.
1933 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1934 Also via indented variable definition in the Makefile.
1935 (Debugging by Oliver Heesakkers).
1937 PP/02 Restore caching of spamd results with expanded spamd_address.
1938 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1940 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1941 Improves build reliability. Fix from: Frank Elsner
1943 NM/02 Fix wide character breakage in the rfc2047 coding
1944 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1946 NM/03 Allow underscore in dnslist lookups
1947 Fixes bug 1026. Patch from Graeme Fowler
1949 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1950 Code patches from Adam Ciarcinski of NetBSD.
1952 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1955 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1956 is logged, to avoid truncation. Patch from John Horne.
1958 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1959 Patch from Jakob Hirsch.
1961 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1962 SQL string expansion failure details.
1963 Patch from Andrey Oktyabrski.
1965 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1966 Patch from Simon Arlott.
1968 PP/09 New lookups functionality failed to compile on old gcc which rejects
1969 extern declarations in function scope.
1970 Patch from Oliver Fleischmann
1972 PP/10 Use sig_atomic_t for flags set from signal handlers.
1973 Check getgroups() return and improve debugging.
1974 Fixed developed for diagnosis in bug 927 (which turned out to be
1977 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1978 Patch from Mark Zealey.
1980 PP/12 Bugzilla 1056: Improved spamd server selection.
1981 Patch from Mark Zealey.
1983 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1984 Based on patch from Heiko Schlittermann.
1986 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1987 Patch from Uwe Doering, sign-off by Michael Haardt.
1989 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1990 variable declaration deep within a block. Bug and patch from
1993 PP/15 lookups-Makefile IRIX compatibility coercion.
1995 PP/16 Make DISABLE_DKIM build knob functional.
1997 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1998 Patch by Simon Arlott
2000 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2001 variable argument macros. Our copy now differs from upstream.
2007 TF/01 Failure to get a lock on a hints database can have serious
2008 consequences so log it to the panic log.
2010 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2011 controlled using the smtp_confirmation log selector.
2013 TF/03 Include the error message when we fail to unlink a spool file.
2015 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2016 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2017 for maintaining out-of-tree patches for some time.
2019 PP/01 Bugzilla 139: Documentation and portability issues.
2020 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2021 Handle per-OS dynamic-module compilation flags.
2023 PP/02 Let /dev/null have normal permissions.
2024 The 4.73 fixes were a little too stringent and complained about the
2025 permissions on /dev/null. Exempt it from some checks.
2026 Reported by Andreas M. Kirchwitz.
2028 PP/03 Report version information for many libraries, including
2029 Exim version information for dynamically loaded libraries. Created
2030 version.h, now support a version extension string for distributors
2031 who patch heavily. Dynamic module ABI change.
2033 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2034 privilege escalation vulnerability whereby the Exim run-time user
2035 can cause root to append content of the attacker's choosing to
2038 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2041 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2042 If dropping privileges for untrusted macros, we disabled normal logging
2043 on the basis that it would fail; for the Exim run-time user, this is not
2044 the case, and it resulted in successful deliveries going unlogged.
2045 Fixed. Reported by Andreas Metzler.
2051 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2052 only prepend for the Resent-* case. Fixes regression introduced in
2053 Exim 4.70 by NM/22 for Bugzilla 607.
2055 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2056 increasing numbers of administrators be bitten by this.
2058 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2060 PP/03 Bugzilla 994: added openssl_options main configuration option.
2062 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2064 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2066 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2068 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2069 without a peer certificate, leading to a segfault because of an
2070 assumption that peers always have certificates. Be a little more
2071 paranoid. Problem reported by Martin Tscholak.
2073 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2074 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2075 NB: ClamAV planning to remove STREAM in "middle of 2010".
2076 CL also introduces -bmalware, various -d+acl logging additions and
2077 more caution in buffer sizes.
2079 PP/09 Implemented reverse_ip expansion operator.
2081 PP/10 Bugzilla 937: provide a "debug" ACL control.
2083 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2085 PP/12 Bugzilla 973: Implement --version.
2087 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2089 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2091 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2093 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2094 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2095 ignore trailing whitespace.
2097 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2099 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2102 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2103 Notification from Dr Andrew Aitchison.
2105 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2106 ExtendedDetectionInfo response format.
2107 Notification from John Horne.
2109 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2112 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2113 XSL and documented dependency on system catalogs, with examples of how
2116 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2119 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2120 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2121 configuration file which is writeable by the Exim user or group.
2123 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2124 of configuration files to cover files specified with the -C option if
2125 they are going to be used with root privileges, not just the default
2128 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2129 option (effectively making it always true).
2131 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2132 files to be used while preserving root privileges.
2134 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2135 that rogue child processes cannot use them.
2137 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2138 run-time user, instead of root.
2140 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2141 Exim run-time user without dropping privileges.
2143 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2144 result string, instead of calling string_vformat() twice with the same
2147 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2148 for other users. Others should always drop root privileges if they use
2149 -C on the command line, even for a whitelisted configure file.
2151 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2153 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2159 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2160 $data_path, and $header_path variables; fixed documentation bugs and
2163 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2164 exipick to access non-standard spools, including the "frozen" queue
2167 NM/01 Bugzilla 965: Support mysql stored procedures.
2168 Patch from Alain Williams
2170 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2172 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2173 Patch from Andreas Metzler
2175 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2176 Patch from Kirill Miazine
2178 NM/05 Bugzilla 671: Added umask to procmail example.
2180 JJ/03 installed exipick 20100323.0, fixing doc bug
2182 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2183 directory. Notification and patch from Dan Rosenberg.
2185 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2187 TK/02 Improve log output when DKIM signing operation fails.
2189 MH/01 Treat the transport option dkim_domain as a colon separated
2190 list, not as a single string, and sign the message with each element,
2191 omitting multiple occurences of the same signer.
2193 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2194 Bugzilla 985, 986. Patch by Simon Arlott
2196 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2197 Patch by Simon Arlott
2199 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2200 MBX locking. Notification from Dan Rosenberg.
2206 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2208 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2210 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2212 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2214 NM/04 Bugzilla 915: Flag broken perl installation during build.
2220 TK/01 Added patch by Johannes Berg that expands the main option
2221 "spamd_address" if it starts with a dollar sign.
2223 TK/02 Write list of recipients to X-Envelope-Sender header when building
2224 the mbox-format spool file for content scanning (suggested by Jakob
2227 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2228 (http://www.dcc-servers.net/) support via dccifd. Activated by
2229 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2231 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2232 by Mark Daniel Reidel <mr@df.eu>.
2234 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2235 When building exim an external PCRE library is now needed -
2236 PCRE is a system library on the majority of modern systems.
2237 See entry on PCRE_LIBS in EDITME file.
2239 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2240 conversation. Added nologin parameter to request.
2241 Patch contributed by Kirill Miazine.
2243 TF/01 Do not log submission mode rewrites if they do not change the address.
2245 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2247 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2248 log files in place. Contributed by Roberto Lima.
2250 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2252 TF/03 Bugzilla 615: When checking the local_parts router precondition
2253 after a local_part_suffix or local_part_prefix option, Exim now
2254 does not use the address's named list lookup cache, since this
2255 contains cached lookups for the whole local part.
2257 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2258 Robert Millan. Documentation is in experimental-spec.txt.
2260 TF/04 Bugzilla 668: Fix parallel build (make -j).
2262 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2264 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2265 Patch provided by Jan Srzednicki.
2267 TF/05 Leading white space used to be stripped from $spam_report which
2268 wrecked the formatting. Now it is preserved.
2270 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2271 that they are available at delivery time.
2273 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2275 TF/08 TLS error reporting now respects the incoming_interface and
2276 incoming_port log selectors.
2278 TF/09 Produce a more useful error message if an SMTP transport's hosts
2279 setting expands to an empty string.
2281 NM/06 Bugzilla 744: EXPN did not work under TLS.
2282 Patch provided by Phil Pennock.
2284 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2285 Patch provided by Richard Godbee.
2287 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2288 acl_smtp_notquit, added index entry.
2290 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2291 Patch provided by Eugene Bujak.
2293 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2294 accept(). Patch provided by Maxim Dounin.
2296 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2297 Patch provided by Phil Pennock.
2299 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2301 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2302 Patch provided by Brad "anomie" Jorsch.
2304 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2305 Patch provided by Dean Brooks.
2307 TK/05 Add native DKIM support (does not depend on external libraries).
2309 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2310 Patch provided by Graeme Fowler.
2312 NM/16 Bugzilla 851: Documentation example syntax fix.
2314 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2316 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2319 NM/19 Bugzilla 745: TLS version reporting.
2320 Patch provided by Phil Pennock.
2322 NM/20 Bugzilla 167: bool: condition support.
2323 Patch provided by Phil Pennock.
2325 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2326 clients. Patch provided by Phil Pennock.
2328 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2329 Patch provided by Brad "anomie" Jorsch.
2331 NM/23 Bugzilla 687: Fix misparses in eximstats.
2332 Patch provided by Heiko Schlittermann.
2334 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2335 Patch provided by Heiko Schlittermann.
2337 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2338 plus update to original patch.
2340 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2342 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2343 Patch provided by David Brownlee.
2345 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2347 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2349 NM/30 Bugzilla 888: TLS documentation bugfixes.
2351 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2353 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2354 Unlike the original bugzilla I have changed all shell scripts in src tree.
2356 NM/33 Bugzilla 898: Transport filter timeout fix.
2357 Patch by Todd Rinaldo.
2359 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2360 Patch by Serge Demonchaux.
2362 NM/35 Bugzilla 39: Base64 decode bug fixes.
2363 Patch by Jakob Hirsch.
2365 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2367 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2369 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2371 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2377 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2378 ALT-N's libdkim that I have put here:
2379 http://duncanthrax.net/exim-experimental/
2381 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2382 were called 'true' and it seems that C99 defines that as a reserved
2383 keyword to be used with 'bool' variable types. That means you could
2384 not include C99-style headers which use bools without triggering
2385 build errors in sieve.c.
2387 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2388 as mailq or other aliases. Changed the --help handling significantly
2389 to do whats expected. exim_usage() emits usage/help information.
2391 SC/01 Added the -bylocaldomain option to eximstats.
2393 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2395 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2397 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2403 PH/01 Another patch from the Sieve maintainer.
2405 PH/02 When an IPv6 address is converted to a string for single-key lookup
2406 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2407 dots are used instead of colons so that keys in lsearch files need not
2408 contain colons. This was done some time before quoting was made available
2409 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2410 (notated using the quote facility) so as to distinguish them from IPv4
2411 keys. This meant that lookups for IP addresses in host lists did not work
2412 for iplsearch lookups.
2414 This has been fixed by arranging for IPv6 addresses to be expressed with
2415 colons if the lookup type is iplsearch. This is not incompatible, because
2416 previously such lookups could never work.
2418 The situation is now rather anomalous, since one *can* have colons in
2419 ordinary lsearch keys. However, making the change in all cases is
2420 incompatible and would probably break a number of configurations.
2422 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2425 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2426 conversion specification without a maximum field width, thereby enabling
2427 a rogue spamd server to cause a buffer overflow. While nobody in their
2428 right mind would setup Exim to query an untrusted spamd server, an
2429 attacker that gains access to a server running spamd could potentially
2430 exploit this vulnerability to run arbitrary code as the Exim user.
2432 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2433 $primary_hostname instead of what libspf2 thinks the hosts name is.
2435 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2436 a directory entry by the name of the lookup key. Previously, if a
2437 symlink pointed to a non-existing file or a file in a directory that
2438 Exim lacked permissions to read, a lookup for a key matching that
2439 symlink would fail. Now it is enough that a matching directory entry
2440 exists, symlink or not. (Bugzilla 503.)
2442 PH/03 The body_linecount and body_zerocount variables are now exported in the
2445 PH/04 Added the $dnslist_matched variable.
2447 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2448 This means they are set thereafter only if the connection becomes
2451 PH/06 Added the client_condition to authenticators so that some can be skipped
2452 by clients under certain conditions.
2454 PH/07 The error message for a badly-placed control=no_multiline_responses left
2455 "_responses" off the end of the name.
2457 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2459 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2460 (without spaces) instead of just copying the configuration text.
2462 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2464 PH/11 Added $max_received_linelength.
2466 PH/12 Added +ignore_defer and +include_defer to host lists.
2468 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2469 way in which PCRE > 7.0 is built.
2471 PH/14 Implemented queue_only_load_latch.
2473 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2474 MAIL command. The effect was to mangle the value on 64-bit systems.
2476 PH/16 Another patch from the Sieve maintainer.
2478 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2480 PH/18 If a system quota error occurred while trying to create the file for
2481 a maildir delivery, the message "Mailbox is full" was not appended to the
2482 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2483 only to a quota excession during the actual writing of the file.
2485 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2486 characters?) which causes problems in log lines. The DN values are now
2487 passed through string_printing() before being added to log lines.
2489 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2490 and InterBase are left for another time.)
2492 PH/21 Added message_body_newlines option.
2494 PH/22 Guard against possible overflow in moan_check_errorcopy().
2496 PH/23 POSIX allows open() to be a macro; guard against that.
2498 PH/24 If the recipient of an error message contained an @ in the local part
2499 (suitably quoted, of course), incorrect values were put in $domain and
2500 $local_part during the evaluation of errors_copy.
2506 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2507 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2510 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2511 issue a MAIL command.
2513 PH/02 In an ACL statement such as
2515 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2517 if a client was not listed at all, or was listed with a value other than
2518 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2519 the condition was not true (as it should be), so access was not denied.
2520 The bug was that the ! inversion was incorrectly passed on to the second
2521 item. This has been fixed.
2523 PH/03 Added additional dnslists conditions == and =& which are different from
2524 = and & when the dns lookup returns more than one IP address.
2526 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2527 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2529 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2530 FSYNC, which compiles an option called disable_fsync that allows for
2531 bypassing fsync(). The documentation is heavily laced with warnings.
2533 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2535 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2536 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2537 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2538 including adding "make clean"; (3) Added -fPIC when compiling the test
2539 dynamically loaded module, to get rid of a warning.
2541 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2542 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2543 = 0s. The bug is otherwise harmless.
2545 PH/07 There was a bug in the dovecot authenticator such that the value of
2546 $auth1 could be overwritten, and so not correctly preserved, after a
2547 successful authentication. This usually meant that the value preserved by
2548 the server_setid option was incorrect.
2550 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2552 PH/09 Installed PCRE release 7.0.
2554 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2555 run for batched SMTP input. It is now run at the start of every message
2556 in the batch. While fixing this I discovered that the process information
2557 (output by running exiwhat) was not always getting set for -bs and -bS
2558 input. This is fixed, and it now also says "batched" for BSMTP.
2560 PH/11 Added control=no_pipelining.
2562 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2563 patch, slightly modified), and move the expansion of helo_data till after
2564 the connection is made in the smtp transport (so it can use these
2567 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2569 PH/14 Added log_selector = +pid.
2571 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2573 PH/16 Add ${if forany and ${if forall.
2575 PH/17 Added dsn_from option to vary the From: line in DSNs.
2577 PH/18 Flush SMTP output before performing a callout, unless control =
2578 no_callout_flush is set.
2580 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2581 was true (the default) a successful delivery failed to delete the retry
2582 item, thus causing premature timeout of the address. The bug is now
2585 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2587 PH/21 Long custom messages for fakedefer and fakereject are now split up
2588 into multiline responses in the same way that messages for "deny" and
2589 other ACL rejections are.
2591 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2592 with slight modification.
2594 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2595 draft, changing the syntax and factoring some duplicate code".
2597 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2598 for deliveries of the second and subsequent messages over the same SMTP
2601 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2602 ${reduce, with only minor "tidies".
2604 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2606 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2607 expansion side effects.
2609 PH/27 When a message times out after an over-quota error from an Exim-imposed
2610 quota, the bounce message says "mailbox is full". This message was not
2611 being given when it was a system quota that was exceeded. It now should
2614 MH/03 Made $recipients available in local_scan(). local_scan() already has
2615 better access to the recipient list through recipients_list[], but
2616 $recipients can be useful in postmaster-provided expansion strings.
2618 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2619 in the case of a MAIL command with additional options following the
2620 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2621 were accidentally chopped off.
2623 PH/29 SMTP synchronization checks are implemented when a command is read -
2624 there is a check that no more input is waiting when there shouldn't be
2625 any. However, for some commands, a delay in an ACL can mean that it is
2626 some time before the response is written. In this time, more input might
2627 arrive, invalidly. So now there are extra checks after an ACL has run for
2628 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2629 pipelining has not been advertised.
2631 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2633 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2634 returned in the SMTP response when smtp_return_error_details was false.
2635 This has been fixed.
2637 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2638 instead of the C I/O that was originally supplied, because problems were
2639 reported on Solaris.
2641 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2642 Exim which did not show up earlier: it was assuming that a call to
2643 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2644 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2645 was a macro that became an assignment, so it seemed to work. This has
2646 changed to a proper function call with a void return, hence the compile
2647 error. Exim's code has been fixed.
2649 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2652 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2654 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2656 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2657 because it turns out to be too big or there is a timeout) there is no
2658 "Completed" line in the log. When some messages of this type were
2659 selected by exigrep, they were listed as "not completed". Others were
2660 picked up by some special patterns. I have improved the selection
2661 criteria to be more general.
2663 PH/38 The host_find_failed option in the manualroute router can now be set
2664 to "ignore", to completely ignore a host whose IP address cannot be
2665 found. If all hosts are ignored, the behaviour is controlled by the new
2666 host_all_ignored option.
2668 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2669 homing or because of multiple MX records with /mx) generated more than
2670 one IP address, and the following item turned out to be the local host,
2671 all the secondary addresses of the first item were incorrectly removed
2672 from the list, along with the local host and any following hosts (which
2673 is what is supposed to happen).
2675 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2676 whoever called Exim into the -H file. In the case of the daemon it was
2677 behaving confusingly. When first started, it used values for whoever
2678 started the daemon, but after a SIGHUP it used the Exim user (because it
2679 calls itself on a restart). I have changed the code so that it now always
2682 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2683 message are rejected with the same error (e.g. no authentication or bad
2684 sender address), and a DATA command is nevertheless sent (as can happen
2685 with PIPELINING or a stupid MUA), the error message that was given to the
2686 RCPT commands is included in the rejection of the DATA command. This is
2687 intended to be helpful for MUAs that show only the final error to their
2690 PH/42 Another patch from the Sieve maintainer.
2692 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2693 Eximstats - Fixed some broken HTML links and added missing column headers
2695 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2696 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2698 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2700 PH/43 Yet another patch from the Sieve maintainer.
2702 PH/44 I found a way to check for a TCP/IP connection going away before sending
2703 the response to the final '.' that terminates a message, but only in the
2704 case where the client has not sent further data following the '.'
2705 (unfortunately, this is allowed). However, in many cases there won't be
2706 any further data because there won't be any more messages to send. A call
2707 to select() can be used: if it shows that the input is "ready", there is
2708 either input waiting, or the socket has been closed. An attempt to read
2709 the next input character can distinguish the two cases. Previously, Exim
2710 would have sent an OK response which the client would never have see.
2711 This could lead to message repetition. This fix should cure that, at
2712 least in a lot of common cases.
2714 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2715 advertised in response to EHLO.
2721 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2722 fixed by 4.65/MH/01 (is this a record?) are fixed:
2724 (i) An empty string was always treated as zero by the numeric comparison
2725 operators. This behaviour has been restored.
2727 (ii) It is documented that the numeric comparison operators always treat
2728 their arguments as decimal numbers. This was broken in that numbers
2729 starting with 0 were being interpreted as octal.
2731 While fixing these problems I realized that there was another issue that
2732 hadn't been noticed. Values of message_size_limit (both the global option
2733 and the transport option) were treated as octal if they started with 0.
2734 The documentation was vague. These values are now always treated as
2735 decimal, and I will make that clear in the documentation.
2741 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2742 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2745 MH/01 Don't check that the operands of numeric comparison operators are
2746 integers when their expansion is in "skipping" mode (fixes bug
2747 introduced by 4.64-PH/07).
2749 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2750 child addresses, Exim now panics and dies. Previously, because the count
2751 is held in a short int, deliveries were likely to be lost. As such a
2752 large number of recipients for a single message is ridiculous
2753 (performance will be very, very poor), I have chosen to impose a limit
2754 rather than extend the field.
2760 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2761 leftover -K file (the existence of which was triggered by #402).
2762 While we were at it, introduced process PID as part of the -K
2763 filename. This should rule out race conditions when creating
2766 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2767 processing considerably. Previous code took too long for large mails,
2768 triggering a timeout which in turn triggers #401.
2770 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2771 in the DK code in transports.c. sendfile() is not really portable,
2772 hence the _LINUX specificness.
2774 TF/01 In the add_headers option to the mail command in an Exim filter,
2775 there was a bug that Exim would claim a syntax error in any
2776 header after the first one which had an odd number of characters
2779 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2780 callout verification, Exim cached a "reject" for the entire domain. This
2781 is correct for most verifications, but it is not correct for a recipient
2782 verification with use_sender or use_postmaster set, because in that case
2783 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2784 case of MAIL FROM:<> rejection from other early rejections (e.g.
2785 rejection of HELO). When verifying a recipient using a non-null MAIL
2786 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2787 Whatever the result of the callout, the value of the domain cache is
2788 left unchanged (for any other kind of callout, getting as far as trying
2789 RCPT means that the domain itself is ok).
2791 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2794 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2795 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2796 session, the daemon ignores ECONNECT errors and logs others; it now
2797 ignores EPIPE as well.
2799 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2800 (quoted-printable decoding).
2802 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2803 later the small subsequent patch to fix an introduced bug.
2805 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2807 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2809 PH/08 An error is now given if message_size_limit is specified negative.
2811 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2812 to be given (somewhat) arbitrary names.
2814 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2817 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2818 miscellaneous code fixes
2820 PH/10 Added the log_reject_target ACL modifier to specify where to log
2823 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2824 hostname. This is wrong, because it relates to the incoming message (and
2825 probably the interface on which it is arriving) and not to the outgoing
2826 callout (which could be using a different interface). This has been
2827 changed to use the value of the helo_data option from the smtp transport
2828 instead - this is what is used when a message is actually being sent. If
2829 there is no remote transport (possible with a router that sets up host
2830 addresses), $smtp_active_hostname is used.
2832 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2833 tweaks were necessary in order to get it to work (see also 21 below):
2834 (a) The code assumed that strncpy() returns a negative number on buffer
2835 overflow, which isn't the case. Replaced with Exim's string_format()
2837 (b) There were several signed/unsigned issues. I just did the minimum
2838 hacking in of casts. There is scope for a larger refactoring.
2839 (c) The code used strcasecmp() which is not a standard C function.
2840 Replaced with Exim's strcmpic() function.
2841 (d) The code set only $1; it now sets $auth1 as well.
2842 (e) A simple test gave the error "authentication client didn't specify
2843 service in request". It would seem that Dovecot has changed its
2844 interface. Fortunately there's a specification; I followed it and
2845 changed what the client sends and it appears to be working now.
2847 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2850 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2851 address A is aliased to B and C, where B exists and C does not. Without
2852 -v the output is "A verified" because verification stops after a
2853 successful redirection if more than one address is generated. However,
2854 with -v the child addresses are also verified. Exim was outputting "A
2855 failed to verify" and then showing the successful verification for C,
2856 with its parentage. It now outputs "B failed to verify", showing B's
2857 parentage before showing the successful verification of C.
2859 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2860 look up a TXT record in a specific list after matching in a combined
2863 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2864 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2865 they consult the DNS. I had assumed they would set it the way they
2866 wanted; and indeed my experiments on Linux seem to show that in some
2867 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2868 To be on the safe side, however, I have now made the interface to
2869 host_find_byname() similar to host_find_bydns(), with an argument
2870 containing the DNS resolver options. The host_find_byname() function now
2871 sets these options at its start, just as host_find_bydns() does. The smtp
2872 transport options dns_qualify_single and dns_search_parents are passed to
2873 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2874 of host_find_byname() use the default settings of RES_DEFNAMES
2875 (qualify_single) but not RES_DNSRCH (search_parents).
2877 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2878 spool_read_header() do less string testing, by means of a preliminary
2879 switch on the second character of optional "-foo" lines. (This is
2880 overdue, caused by the large number of possibilities that now exist.
2881 Originally there were few.) While I was there, I also converted the
2882 str(n)cmp tests so they don't re-test the leading "-" and the first
2883 character, in the hope this might squeeze out yet more improvement.
2885 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2886 flag allowing group syntax was set by the header_syntax check but not
2887 turned off, possible causing trouble later; (2) The flag was not being
2888 set at all for the header_verify test, causing "group"-style headers to
2889 be rejected. I have now set it in this case, and also caused header_
2890 verify to ignore an empty address taken from a group. While doing this, I
2891 came across some other cases where the code for allowing group syntax
2892 while scanning a header line wasn't quite right (mostly, not resetting
2893 the flag correctly in the right place). These bugs could have caused
2894 trouble for malformed header lines. I hope it is now all correct.
2896 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2897 with the "reply" argument non-NULL. The code, however (which originally
2898 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2899 but it didn't always do it. This confused somebody who was copying the
2900 code for some other use. I have removed all the tests.
2902 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2903 feature that was used to support insecure browsers during the U.S. crypto
2904 embargo. It requires special client support, and Exim is probably the
2905 only MTA that supported it -- and would never use it because real RSA is
2906 always available. This code has been removed, because it had the bad
2907 effect of slowing Exim down by computing (never used) parameters for the
2908 RSA_EXPORT functionality.
2910 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2911 authenticator to fail if there's a tab character in the incoming data
2912 (there should never be unless someone is messing about, as it's supposed
2913 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2914 if the connection is using TLS or if the remote IP is the same as the
2915 local IP, and the "valid-client-cert option" if a client certificate has
2918 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2919 authenticators. This can be used for authorization after authentication
2920 succeeds. (In the case of plaintext, it servers for both authentication
2923 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2924 if any retry times were supplied.
2926 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2927 connection, where there is no client IP address to check. In this
2928 situation, the verify now always succeeds.
2930 PH/25 Applied John Jetmore's -Mset patch.
2932 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2934 PH/27 In a string expansion for a processed (not raw) header when multiple
2935 headers of the same name were present, leading whitespace was being
2936 removed from all of them, but trailing whitespace was being removed only
2937 from the last one. Now trailing whitespace is removed from each header
2938 before concatenation. Completely empty headers in a concatenation (as
2939 before) are ignored.
2941 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2942 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2944 PH/29 [Removed. This was a change that I later backed out, and forgot to
2945 correct the ChangeLog entry (that I had efficiently created) before
2946 committing the later change.]
2948 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2949 address errors (4xx response to RCPT) over the same connection as other
2950 messages routed to the same hosts. Such deliveries are always "forced",
2951 so retry times are not inspected. This resulted in far too many retries
2952 for the affected addresses. The effect occurred only when there were more
2953 hosts than the hosts_max_try setting in the smtp transport when it had
2954 the 4xx errors. Those hosts that it had tried were not added to the list
2955 of hosts for which the message was waiting, so if all were tried, there
2956 was no problem. Two fixes have been applied:
2958 (i) If there are any address or message errors in an SMTP delivery, none
2959 of the hosts (tried or untried) are now added to the list of hosts
2960 for which the message is waiting, so the message should not be a
2961 candidate for sending over the same connection that was used for a
2962 successful delivery of some other message. This seems entirely
2963 reasonable: after all the message is NOT "waiting for some host".
2964 This is so "obvious" that I'm not sure why it wasn't done
2965 previously. Hope I haven't missed anything, but it can't do any
2966 harm, as the worst effect is to miss an optimization.
2968 (ii) If, despite (i), such a delivery is accidentally attempted, the
2969 routing retry time is respected, so at least it doesn't keep
2970 hammering the server.
2972 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2973 in ${readsocket because some servers need this prod.
2975 PH/32 Added some extra debug output when updating a wait-xxx database.
2977 PH/33 The hint "could be header name not terminated by colon", which has been
2978 given for certain expansion errors for a long time, was not being given
2979 for the ${if def:h_colon_omitted{... case.
2981 PH/34 The spec says: "With one important exception, whenever a domain list is
2982 being scanned, $domain contains the subject domain." There was at least
2983 one case where this was not true.
2985 PH/35 The error "getsockname() failed: connection reset by peer" was being
2986 written to the panic log as well as the main log, but it isn't really
2987 panic-worthy as it just means the connection died rather early on. I have
2988 removed the panic log writing for the ECONNRESET error when getsockname()
2991 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2992 runs only) independently of the message's sender address. This meant
2993 that, if the 4xx error was in fact related to the sender, a different
2994 message to the same recipient with a different sender could confuse
2995 things. In particular, this can happen when sending to a greylisting
2996 server, but other circumstances could also provoke similar problems.
2997 I have changed the default so that the retry time for these errors is now
2998 based a combination of the sender and recipient addresses. This change
2999 can be overridden by setting address_retry_include_sender=false in the
3002 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3003 remote server are returned as part of bounce messages. This was not
3004 happening for LMTP over a pipe (the lmtp transport), but now it is the
3005 same for both kinds of LMTP.
3007 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3008 in header lines that were in fact CNAMEs. This is no longer the case.
3010 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3011 and queue runs started by the daemon processed all messages. This has
3012 been fixed so that -R and -S can now usefully be given with -q<time>.
3014 PH/40 Import PCRE release 6.7 (fixes some bugs).
3016 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3018 PH/42 Give an error if -q is specified more than once.
3020 PH/43 Renamed the variables $interface_address and $interface_port as
3021 $received_ip_address and $received_port, to make it clear that these
3022 values apply to message reception, and not to the outgoing interface when
3023 a message is delivered. (The old names remain recognized, of course.)
3025 PH/44 There was no timeout on the connect() call when using a Unix domain
3026 socket in the ${readsocket expansion. There now is.
3028 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3029 be meaningful with "accept".
3031 SC/01 Eximstats V1.43
3032 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3034 SC/02 Eximstats V1.44
3035 Use a glob alias rather than an array ref in the generated
3036 parser. This improves both readability and performance.
3038 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3039 Collect SpamAssassin and rejection statistics.
3040 Don't display local sender or destination tables unless
3041 there is data to show.
3042 Added average volumes into the top table text output.
3044 SC/04 Eximstats V1.46
3045 Collect data on the number of addresses (recipients)
3046 as well as the number of messages.
3048 SC/05 Eximstats V1.47
3049 Added 'Message too big' to the list of mail rejection
3050 reasons (thanks to Marco Gaiarin).
3052 SC/06 Eximstats V1.48
3053 Mainlog lines which have GMT offsets and are too short to
3054 have a flag are now skipped.
3056 SC/07 Eximstats V1.49 (Alain Williams)
3057 Added the -emptyok flag.
3059 SC/08 Eximstats V1.50
3060 Fixes for obtaining the IP address from reject messages.
3062 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3063 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3064 whitespace changes from 4.64-PH/27
3066 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3069 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3070 are found, allow negative numbers in numeric criteria)
3072 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3074 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3075 to match changes made in 4.64-PH/43
3077 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3079 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3080 "domain" is actually an IP address, and force a failure. This locks out
3081 those revolvers/nameservers that support "A-for-A" lookups, in
3082 contravention of the specifications.
3084 PH/48 When a host name was looked up from an IP address, and the subsequent
3085 forward lookup of the name timed out, the host name was left in
3086 $sender_host_name, contrary to the specification.
3088 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3089 restricted to single-key lookups, Exim was not diagnosing an error if
3090 * or *@ was used with a query-style lookup.
3092 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3094 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3095 long ago, but noone interested enough thought of it. Let's just say that
3096 the "1.1" means that there are some new functions that weren't there at
3097 some point in the past.
3099 PH/51 Error processing for expansion failure of helo_data from an smtp
3100 transport during callout processing was broken.
3102 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3103 tested/used via the -bh/-bhc/-bs options.
3105 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3106 bug, fixed in subsequent PCRE releases).
3108 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3109 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3111 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3117 SC/01 Use a glob alias rather than an array ref in eximstats generated
3118 parser. This improves both readability and performance.
3120 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3121 Don't display local sender or destination tables in eximstats unless
3122 there is data to show.
3123 Added average volumes into the eximstats top table text output.
3125 SC/03 Collect data on the number of addresses (recipients) as well
3126 as the number of messages in eximstats.
3128 TF/01 Correct an error in the documentation for the redirect router. Exim
3129 does (usually) call initgroups() when daemonizing.
3131 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3132 with consistent privilege compared to when running as a daemon.
3134 TF/03 Note in the spec that $authenticated_id is not set for local
3135 submissions from trusted users.
3137 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3138 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3140 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3141 by adding some example configuration directives to the default
3142 configuration file. A little bit of work is required to uncomment the
3143 directives and define how usernames and passwords are checked, but
3144 there is now a framework to start from.
3146 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3147 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3148 without this. I don't know how relevant this is to other LDAP libraries.
3150 PH/02 Add the verb name to the "unknown ACL verb" error.
3152 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3154 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3156 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3157 directory not expanded when it should be if an expanded home directory
3158 was set for the address (which is overridden by the transport).
3160 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3163 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3164 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3165 because it is too late at that time, and has no effect.
3167 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3168 security issue with \' (bugzilla #107). I could not use the
3169 PQescapeStringConn() function, because it needs a PGconn value as one of
3172 PH/08 When testing addresses using -bt, indicate those final addresses that
3173 are duplicates that would not cause an additional delivery. At least one
3174 person was confused, thinking that -bt output corresponded to deliveries.
3175 (Suppressing duplicates isn't a good idea as you lose the information
3176 about possibly different redirections that led to the duplicates.)
3178 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3179 systems where poll() doesn't work, in particular OS X.
3181 PH/10 Added more information to debugging output for retry time not reached.
3183 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3184 operations in malware.c.
3186 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3189 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3190 syslog_duplication set false, log lines that would normally be written
3191 both the the main log and to the reject log were not written to syslog at
3194 PH/14 In the default configuration, change the use of "message" in ACL warn
3195 statements to "add_header".
3197 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3198 not followed by a command (e.g. "seen endif").
3200 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3201 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3204 PH/17 Added extra conditions to the default value of delay_warning_condition
3208 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3209 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3210 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3213 The Auto-Submitted: and various List- headers are standardised, whereas I
3214 don't think Precedence: ever was.
3216 PH/18 Refactored debugging code in route_finduser() to show more information,
3217 in particular, the error code if getpwnam() issues one.
3219 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3220 This is apparently needed in addition to the PH/07 change above to avoid
3221 any possible encoding problems.
3223 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3224 but not after initializing Perl.
3226 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3227 output them only if debugging. By default they are written stderr,
3228 apparently, which is not desirable.
3230 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3233 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3236 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3238 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3239 authenticated or an ident call has been made. Suppress the default
3240 values for $authenticated_id and $authenticated_sender (but permit -oMai
3241 and -oMas) when testing with -bh.
3243 PH/24 Re-jigged the order of the tests in the default configuration so that the
3244 tests for valid domains and recipients precede the DNS black list and CSA
3245 tests, on the grounds that those ones are more expensive.
3247 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3248 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3249 This bug exists in every version of Exim that I still have, right back to
3252 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3253 However, an attempt to turn on case-sensitivity in a regex key by
3254 including (?-i) didn't work because the subject string was already
3255 lowercased, and the effects were non-intuitive. It turns out that a
3256 one-line patch can be used to allow (?-i) to work as expected.
3262 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3263 other effects) broke the use of negated acl sub-conditions.
3265 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3268 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3269 "Deny" causes Exim to reject the incoming connection with a 554 error.
3270 Unfortunately, if there is a major crisis, such as a disk failure,
3271 tcp-wrappers gives "deny", whereas what one would like would be some
3272 kind of temporary error. A kludge has been added to help with this.
3273 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3274 554 error is used if errno is still zero or contains ENOENT (which occurs
3275 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3278 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3280 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3281 errors. Otherwise a message that provokes a temporary error (when other
3282 messages do not) can cause a whole host to time out.
3284 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3285 addresses were routed directly to files or pipes from a redirect router.
3286 File deliveries just didn't batch; pipe deliveries might have suffered
3289 PH/06 A failure to get a lock for a hints database would erroneously always say
3290 "Failed to get write lock", even when it was really a read lock.
3292 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3293 of 0600. This has been changed to use the value of the lockfile_mode
3294 option (which defaults to 0600).
3296 PH/08 Applied small patch from the Sieve maintainer.
3298 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3299 folder from quota calculations, a direct delivery into this folder messed
3300 up the contents of the maildirsize file. This was because the regex was
3301 used only to exclude .Trash (or whatever) when the size of the mailbox
3302 was calculated. There was no check that a delivery was happening into an
3303 excluded directory. This bug has been fixed by ignoring all quota
3304 processing for deliveries into excluded directories.
3306 PH/10 Added the maildirfolder_create_regex option to appendfile.
3312 PH/01 The code for finding all the local interface addresses on a FreeBSD
3313 system running IPv6 was broken. This may well have applied to all BSD
3314 systems, as well as to others that have similar system calls. The broken
3315 code found IPv4 interfaces correctly, but gave incorrect values for the
3316 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3317 that it would not match correctly against @[] and not recognize the IPv6
3320 PH/02 The ipliteral router was not recognizing addresses of the form user@
3321 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3323 PH/03 Added disable_ipv6.
3325 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3326 decoded form, because it is most often used to construct To: headers
3327 lines in autoreplies, and the decoded form may well be syntactically
3328 invalid. However, $reply_address has leading white space removed, and all
3329 newlines turned into spaces so that the autoreply transport does not
3332 PH/05 If group was specified without a user on a router, and no group or user
3333 was specified on a transport, the group from the router was ignored.
3335 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3336 for visible compile-time settings that can be used to change these
3337 numbers, for those that want even more. Backwards compatibility with old
3338 spool files has been maintained. However, going back to a previous Exim
3339 release will lost any variables that are in spool files.
3341 PH/07 Two small changes when running in the test harness: increase delay when
3342 passing a TCP/IP connection to a new process, in case the original
3343 process has to generate a bounce, and remove special handling of
3344 127.0.0.2 (sic), which is no longer necessary.
3346 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3347 be the same on different OS.
3349 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3352 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3353 whether --show-vars was specified or not
3355 JJ/02 exipick: Added support for new ACL variable spool format introduced
3358 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3359 syntactically invalid From: or Reply-to: line, and a filter used this to
3360 generate an autoreply, and therefore failed to obtain an address for the
3361 autoreply, Exim could try to deliver to a non-existent relative file
3362 name, causing unrelated and misleading errors. What now happens is that
3363 it logs this as a hard delivery error, but does not attempt to create a
3366 PH/11 The exinext utility has a -C option for testing purposes, but although
3367 the given file was scanned by exinext itself; it wasn't being passed on
3368 when Exim was called.
3370 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3371 an end-of-file indication when reading a command response.
3373 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3374 compiled. In many other places in Exim, IPv6 addresses are always
3375 recognized, so I have changed this. It also means that IPv4 domain
3376 literals of the form [IPV4:n.n.n.n] are now always recognized.
3378 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3379 used if the router is not running as root, for example, when verifying at
3380 ACL time, or when using -bh. The debugging output from this situation was
3381 non-existent - all you got was a failure to exec. I have made two
3384 (a) Failures to set uid/gid, the current directory, or a process leader
3385 in a subprocess such as that created by queryprogram now generate
3386 suitable debugging output when -d is set.
3388 (b) The queryprogram router detects when it is not running as root,
3389 outputs suitable debugging information if -d is set, and then runs
3390 the subprocess without attempting to change uid/gid.
3392 PH/15 Minor change to Makefile for building test_host (undocumented testing
3395 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3396 additional section of a DNS packet that returns MX or SRV records.
3397 Instead, it always explicitly searches for A/AAAA records. This avoids
3398 major problems that occur when a DNS server includes only records of one
3399 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3400 fixed another bug: if SRV records were looked up and the corresponding
3401 address records were *not* found in the additional section, the port
3402 values from the SRV records were lost.
3404 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3405 using the correct key (the original address) when searching the retry
3406 rules in order to find which one to use for generating the retry hint.
3408 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3409 adding the default one. Similarly, if it contains a Reply-To: header, the
3410 errors_reply_to option, if set, is not used.
3412 PH/19 When calculating a retry time, Exim used to measure the "time since
3413 failure" by looking at the "first failed" field in the retry record. Now
3414 it does not use this if it is later than than the arrival time of the
3415 message. Instead it uses the arrival time. This makes for better
3416 behaviour in cases where some deliveries succeed, thus re-setting the
3417 "first failed" field. An example is a quota failure for a huge message
3418 when small messages continue to be delivered. Without this change, the
3419 "time since failure" will always be short, possible causing more frequent
3420 delivery attempts for the huge message than are intended.
3421 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3423 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3424 $1, $2, $3) because the numerical variables can be reset during some
3425 expansion items (e.g. "match"), thereby losing the authentication data.
3427 PH/21 Make -bV show the size of off_t variables so that the test suite can
3428 decide whether to run tests for quotas > 2G.
3430 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3431 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3432 filecount value is greater than 2G or if a quota value is greater than 2G
3433 on a system where the size of off_t is not greater than 4, a panic error
3436 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3437 never match. The debug and -bh output now contains an explicit error
3438 message indicating a malformed IPv4 address or mask.
3440 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3441 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3442 PH/23 above applies.
3444 PH/25 Do not write to syslog when running in the test harness. The only
3445 occasion when this arises is a failure to open the main or panic logs
3446 (for which there is an explicit test).
3448 PH/26 Added the /no_tell option to "control=freeze".
3450 PH/27 If a host name lookup failed very early in a connection, for example, if
3451 the IP address matched host_lookup and the reverse lookup yielded a name
3452 that did not have a forward lookup, an error message of the form "no IP
3453 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3454 could be logged. Now it outputs the IP address instead of "NULL".
3456 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3457 allows the sender and the authenticated sender to be set when
3458 submitting a message from within Exim. Since child_open_exim() is
3459 documented for local_scan(), the new function should be too.
3461 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3462 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3463 results in an empty string is now treated as unset.
3465 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3467 PH/31 Added sender_verify_fail logging option.
3469 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3470 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3471 tidied the source and removed it altogether.
3473 PH/33 When a queue run was abandoned because the load average was too high, a
3474 log line was always written; now it is written only if the queue_run log
3475 selector is set. In addition, the log line for abandonment now contains
3476 information about the queue run such as the pid. This is always present
3477 in "start" and "stop" lines but was omitted from the "abandon" line.
3479 PH/34 Omit spaces between a header name and the colon in the error message that
3480 is given when verify = headers_syntax fails (if there are lots of them,
3481 the message gets confusing).
3483 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3484 names, as there are now some PTR records that contain slashes. This check
3485 is only to protect against broken name servers that fall over on strange
3486 characters, so the fact that it applies to all lookups doesn't matter.
3488 PH/36 Now that the new test suite is complete, we can remove some of the
3489 special code in Exim that was needed for the old test suite. For example,
3490 sorting DNS records because real resolvers return them in an arbitrary
3491 order. The new test suite's fake resolver always returns records in the
3494 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3495 bounces) except when queue_only is set, to avoid logging races between
3496 the different processes.
3498 PH/38 Panic-die if .include specifies a non-absolute path.
3500 PH/39 A tweak to the "H" retry rule from its user.
3502 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3503 a label. They prevented compilation on older perls.
3505 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3506 a warning to be raised on newish perls.
3508 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3509 on queue. Changes to match documented behaviour of showing count of
3510 messages matching specified criteria.
3512 PH/40 Changed the default ident timeout from 30s to 5s.
3514 PH/41 Added support for the use of login_cap features, on those BSD systems
3515 that have them, for controlling the resources used by pipe deliveries.
3517 PH/42 The content-scanning code uses fopen() to create files in which to put
3518 message data. Previously it was not paying any attention to the mode of
3519 the files. Exim runs with umask(0) because the rest of the code creates
3520 files with open(), and sets the required mode explicitly. Thus, these
3521 files were ending up world-writeable. This was not a big issue, because,
3522 being within the spool directory, they were not world-accessible. I have
3523 created a function called modefopen, which takes an additional mode
3524 argument. It sets umask(777), creates the file, chmods it to the required
3525 mode, then resets the umask. All the relevant calls to fopen() in the
3526 content scanning code have been changed to use this function.
3528 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3529 to 24 hours. This avoids potential overflow problems when processing G
3530 and H retry rules. I suspect nobody ever tinkers with this value.
3532 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3534 PH/45 When the plaintext authenticator is running as a client, the server's
3535 challenges are checked to ensure they are valid base64 strings. By
3536 default, the authentication attempt is cancelled if an invalid string is
3537 received. Setting client_ignore_invalid_base64 true ignores these errors.
3538 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3539 they are received. Thus, the responses can be made to depend on the
3540 challenges. If an invalid string is ignored, an empty string is placed in
3543 PH/46 Messages that are created by the autoreply transport now contains a
3544 References: header, in accordance with RFCs 2822 and 3834.
3546 PH/47 Added authenticated_sender_force to the smtp transport.
3548 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3550 PH/49 Installed latest patch from the Sieve maintainer.
3552 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3553 was also set, the appendfile transport was unnecessarily scanning a
3554 directory of message files (e.g. for maildir delivery) to find the count
3555 of files (along with the size), even though it did not need this
3556 information. It now does the scan only if it needs to find either the
3557 size of the count of files.
3559 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3561 PH/52 Two bugs concerned with error handling when the smtp transport is
3564 (i) Exim was not creating retry information for temporary errors given
3565 for individual recipients after the DATA command when the smtp transport
3566 was used in LMTP mode. This meant that they could be retried too
3567 frequently, and not timed out correctly.
3569 (ii) Exim was setting the flag that allows error details to be returned
3570 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3571 recipients that were returned after the DATA command.
3573 PH/53 This is related to PH/52, but is more general: for any failing address,
3574 when detailed error information was permitted to be returned to the
3575 sender, but the error was temporary, then after the final timeout, only
3576 "retry timeout exceeded" was returned. Now it returns the full error as
3577 well as "retry timeout exceeded".
3579 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3580 do this, and (what is worse) MTAs that accept it.
3582 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3583 will now be deprecated.
3585 PH/56 New os.c-cygwin from the Cygwin maintainer.
3587 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3588 formats (previously only available in exim formats via -bpr, -bpru,
3589 and -bpra. Now also available in native and exiqgrep formats)
3591 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3592 with very large, slow to parse queues
3594 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3596 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3598 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3599 responses to authentication challenges, though it was showing the
3600 challenges; (ii) I've removed the CR characters from the debug output for
3603 PH/58 Allow for the insertion of a newline as well as a space when a string
3604 is turned into more than one encoded-word during RFC 2047 encoding. The
3605 Sieve code now uses this.
3607 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3608 data_4xx, lost_connection, tls_required.
3610 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3611 message was being sent as an SMTP response.
3613 PH/61 Add -l and -k options to exicyclog.
3615 PH/62 When verifying, if an address was redirected to one new address, so that
3616 verification continued, and the new address failed or deferred after
3617 having set something in $address_data, the value of $address_data was not
3618 passed back to the ACL. This was different to the case when no
3619 redirection occurred. The value is now passed back in both cases.
3621 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3622 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3623 use login_cap.h, so on its own it isn't the distinguishing feature. The
3624 new name refers directly to the setclassresources() function.
3626 PH/65 Added configuration files for NetBSD3.
3628 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3630 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3631 is preferred over IPv4.
3633 PH/68 The bounce_return_message and bounce_return_body options were not being
3634 honoured for bounces generated during the reception of non-SMTP messages.
3635 In particular, this applied to messages rejected by the ACL. This bug has
3636 been fixed. However, if bounce_return_message is true and bounce_return_
3637 body is false, the headers that are returned for a non-SMTP message
3638 include only those that have been read before the error was detected.
3639 (In the case of an ACL rejection, they have all been read.)
3641 PH/69 The HTML version of the specification is now built in a directory called
3642 spec_html instead of spec.html, because the latter looks like a path with
3643 a MIME-type, and this confuses some software.
3645 PH/70 Catch two compiler warnings in sieve.c.
3647 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3648 function verify_get_ident() calls ip_connect() to connect a socket, but
3649 if the "connect()" function timed out, ip_connect() used to close the
3650 socket. However, verify_get_ident() also closes the socket later, and in
3651 between Exim writes to the log, which may get opened at this point. When
3652 the socket was closed in ip_connect(), the log could get the same file
3653 descriptor number as the socket. This naturally causes chaos. The fix is
3654 not to close the socket in ip_connect(); the socket should be closed by
3655 the function that creates it. There was only one place in the code where
3656 this was missing, in the iplookup router, which I don't think anybody now
3657 uses, but I've fixed it anyway.
3659 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3660 well as to direct DNS lookups. Otherwise the handling of names in host
3661 lists is inconsistent and therefore confusing.
3667 PH/01 Two changes to the default runtime configuration:
3669 (1) Move the checks for relay_from_hosts and authenticated clients from
3670 after to before the (commented out) DNS black list checks.
3672 (2) Add control=submission to the relay_from_hosts and authenticated
3673 clients checks, on the grounds that messages accepted by these
3674 statements are most likely to be submissions.
3676 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3678 (1) Generate an error if the third argument for the ${prvs expansion is
3681 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3684 (3) Reset the variables that are obtained from the first argument of
3685 ${prvscheck and used in the second argument before leaving the code,
3686 because their memory is reclaimed, so using them afterwards may do
3689 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3690 one (it's much easier than Tom thought :-).
3692 (5) Because of (4), we can now allow for the use of $prvscheck_result
3693 inside the third argument.
3695 PH/03 For some reason, the default setting of PATH when running a command from
3696 a pipe transport was just "/usr/bin". I have changed it to
3699 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3700 anything to be listed in the output from -bV.
3702 PH/05 When a filter generated an autoreply, the entire To: header line was
3703 quoted in the delivery log line, like this:
3705 => >A.N.Other <ano@some.domain> <original@ddress> ...
3707 This has been changed so that it extracts the operative address. There
3708 may be more than one such address. If so, they are comma-separated, like
3711 => >ano@some.domain,ona@other.domain <original@ddress> ...
3713 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3714 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3715 not being looked up in the rDNS to get a host name, Exim was showing the
3716 IP address twice in Received: lines, even though the IP addresses were
3717 identical. For example:
3719 Received: from [1.2.3.4] (helo=[1.2.3.4])
3721 However, if the real host name was known, it was omitting the HELO data
3722 if it matched the actual IP address. This has been tidied up so that it
3723 doesn't show the same IP address twice.
3725 PH/07 When both +timestamp and +memory debugging was on, the value given by
3726 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3727 called by the debug printing, thereby overwriting the timestamp buffer.
3728 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3730 PH/08 When the original message was included in an autoreply transport, it
3731 always said "this is a copy of the message, including all the headers",
3732 even if body_only or headers_only was set. It now gives an appropriate
3735 PH/09 Applied a patch from the Sieve maintainer which:
3737 o fixes some comments
3738 o adds the (disabled) notify extension core
3739 o adds some debug output for the result of if/elsif tests
3740 o points to the current vacation draft in the documentation
3741 and documents the missing references header update
3745 o fixes a bug in processing the envelope test (when testing
3746 multiple envelope elements, the last element determined the
3749 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3750 Electronic Mail") by including:
3752 Auto-submitted: auto-generated
3754 in the messages that it generates (bounce messages and others, such as
3755 warnings). In the case of bounce messages for non-SMTP messages, there was
3756 also a typo: it was using "Auto_submitted" (underscore instead of
3757 hyphen). Since every message generated by Exim is necessarily in response
3758 to another message, thes have all been changed to:
3760 Auto-Submitted: auto-replied
3762 in accordance with these statements in the RFC:
3764 The auto-replied keyword:
3766 - SHOULD be used on messages sent in direct response to another
3767 message by an automatic process,
3769 - MUST NOT be used on manually-generated messages,
3771 - MAY be used on Delivery Status Notifications (DSNs) and Message
3772 Disposition Notifications (MDNs),
3774 - MUST NOT be used on messages generated by automatic or periodic
3775 processes, except for messages which are automatic responses to
3778 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3779 to the default Received: header definition.
3781 PH/12 Added log selector acl_warn_skipped (default on).
3783 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3784 variables because (a) they are in the wrong storage pool and (b) even if
3785 they were copied, it wouldn't work properly because of the caching.
3787 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3788 checking when decoding. Apparently there are clients that generate
3789 overlong encoded strings. Why am I not surprised?
3791 PH/15 If the first argument of "${if match_address" was not empty, but did not
3792 contain an "@" character, Exim crashed. Now it writes a panic log message
3793 and treats the condition as false.
3795 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3797 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3798 extension "envelope-auth". The code is finished and in agreement with
3799 other implementations, but there is no documentation so far and in fact,
3800 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3801 not changing the active code.
3803 Print executed "if" and "elsif" statements when debugging is used. This
3804 helps a great deal to understand what a filter does.
3806 Document more things not specified clearly in RFC3028. I had all this
3807 sorted out, when out of a sudden new issues came to my mind. Oops."
3809 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3812 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3813 canonical form (as documented). However, after a host name lookup from
3814 the IP address, check_host() was doing a simple string comparison with
3815 addresses acquired from the DNS when checking that the found name did
3816 have the original IP as one of its addresses. Since any found IPv6
3817 addresses are likely to be in abbreviated form, the comparison could
3818 fail. Luckily, there already exists a function for doing the comparison
3819 by converting both addresses to binary, so now that is used instead of
3820 the text comparison.
3822 PH/20 There was another similar case to PH/19, when a complete host name was
3823 given in a host list; looking up its IP address could give an abbreviated
3824 form, whereas the current host's name might or might not be abbreviated.
3825 The same fix has been applied.
3831 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3832 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3835 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3836 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3838 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3840 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3841 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3842 Also, for "Auto-Submitted", treat anything other than "no" as
3843 non-personal, in accordance with RFC 3834. (Previously it treated
3844 anything starting "auto-" as non-personal.)
3846 TF/01 The control=submission/name=... option had a problem with syntax
3847 errors if the name included a slash character. The /name= option
3848 now slurps the rest of the string, so it can include any characters
3849 but it must come last in the list of options (after /sender_retain
3852 PH/05 Some modifications to the interface to the fake nameserver for the new
3860 TK/01 Added the "success_on_redirect" address verification option. See
3861 NewStuff for rationale and an example.
3863 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3865 PH/02 Patch to exigrep to allow it to work on syslog lines.
3867 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3868 fread() to skip over the body file's header line, because in Cygwin the
3869 header line is locked and is inaccessible.
3871 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3872 co-exist for some time) to make it clear that it is the Exim ID that is
3873 referenced, not the Message-ID: header line.
3875 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3876 string_format() function, because snprintf() does not exist on all
3879 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3880 ${stat: expansion item.
3882 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3883 protocol synchronization error", to keep the pedants happy.
3885 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3886 well as for IRIX systems, when gcc is being used. See the host.c source
3889 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3891 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3894 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3895 transport and to the smtp transport in LMTP mode.
3897 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3899 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3900 run a filter in a subprocess. This could lead to confusion in subsequent
3901 lookups in the parent process. There should also be a search_tidyup() at
3902 the end of the subprocess.
3904 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3905 only if the host matched helo_try_verify_hosts, which caused the
3906 verification to occur when the EHLO/HELO command was issued. The ACL just
3907 tested the remembered result. Now, if a previous verification attempt has
3908 not happened, "verify = helo" does it there and then.
3910 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3912 TK/03 Fix log output including CR from clamd.
3914 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3915 exist provoked a memory error which could cause a segfault.
3917 PH/15 Installed PCRE 6.2
3919 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3921 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3922 of the problem. Specifically, suggested +O2 rather than +O1 for the
3925 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3927 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3928 record, the port was not correctly logged when the outgoing_port log
3929 selector was set (it logged the transort's default port).
3931 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3932 fallback_hosts, and "hosts" in the smtp transport.
3934 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3935 host errors such as "Connection refused".
3937 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3938 authentication with radiusclient 0.4.9:
3940 - Error returned from rc_read_config was caught wrongly
3941 - Username/password not passed on to radius server due to wrong length.
3943 The presumption is that some radiusclient API changes for 4.51/PH/17
3944 were not taken care of correctly. The code is still untested by me (my
3945 Linux distribution still has 0.3.2 of radiusclient), but it was
3946 contributed by a Radius user.
3948 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3949 expanding the "port" option of the smtp transport.
3951 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3952 while reading a MIME header. Thanks to Tom Hughes for a patch.
3954 PH/24 Include config.h inside local_scan.h so that configuration settings are
3957 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3958 that in an ACL for RCPT (for example), you can examine exactly what was
3961 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3962 commands, but it was not correctly comparing the address with the actual
3963 client host address. Thus, it would show the EHLO address in Received:
3964 header lines when this was not necessary.
3966 PH/27 Added the % operator to ${eval:}.
3968 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3969 it should be ignoring failures (because with -C, for example, it has lost
3970 privilege). It wasn't ignoring creation failures other than "already
3973 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3976 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3977 because an input line was too long, either on its own, or by virtue of
3978 too many continuations, the temporary file was not being removed, and the
3979 return code was incorrect.
3981 PH/31 Missing "BOOL" in function definition in filtertest.c.
3983 PH/32 Applied Sieve patches from the maintainer.
3985 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3987 PH/33 Added "verify = not_blind".
3989 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3990 Local/Makefile (with some defaults set). These are used in built scripts
3991 such as exicyclog, but they have never been used in the exim_install
3992 script (though there are many overriding facilities there). I have
3993 arranged that the exim_install script now takes note of these two
3996 PH/35 Installed configuration files for Dragonfly.
3998 PH/36 When a locally submitted message by a trusted user did not contain a
3999 From: header, and the sender address was obtained from -f or from an SMTP
4000 MAIL command, and the trusted user did not use -F to supply a sender
4001 name, $originator_name was incorrectly used when constructing a From:
4002 header. Furthermore, $originator_name was used for submission mode
4003 messages from external hosts without From: headers in a similar way,
4004 which is clearly wrong.
4006 PH/37 Added control=suppress_local_fixups.
4008 PH/38 When log_selector = +received_sender was set, and the addition of the
4009 sender made the log line's construction buffer exactly full, or one byte
4010 less than full, an overflow happened when the terminating "\n" was
4013 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4014 when the result of a list match is failure because a DNS lookup failed.
4016 PH/40 RM_COMMAND is now used in the building process.
4018 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4019 the "build-* directories that it finds.
4021 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4022 address in a domain literal was a prefix of an interface address.
4024 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4025 when verifying a sender address, unless rewrite_headers is false.
4027 PH/44 Wrote a long comment about why errors_to addresses are verified as
4028 recipients, not senders.
4030 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4031 the ratelimit ACL was added.
4033 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4035 PH/46 Added extra information about PostgreSQL errors to the error string.
4037 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4038 avoiding the need to install special zones in a real server. This is
4039 backwards compatible; if it can't find the fake resolver, it drops back.
4040 Thus, both old and new test suites can be run.
4042 TF/02 Added util/ratelimit.pl
4044 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4045 clock is set back in time.
4047 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4048 Candler <B.Candler@pobox.com>.
4050 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4051 OK for addresses that are the result of redirections.
4053 PH/48 A number of further additions for the benefit of the new test suite,
4054 including a fake gethostbyname() that interfaces to the fake DNS resolver
4057 TF/06 The fix for widen_domains has also been applied to qualify_single and
4058 search_parents which are the other dnslookup options that can cause
4061 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4064 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4066 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4067 service" if it encountered an MX record with an empty target hostname.
4068 The message is now "an MX or SRV record indicated no SMTP service".
4070 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4071 if the DNS of the sending site is misconfigured. This is quite a
4072 common situation. This change restores the behaviour of treating a
4073 helo verification defer as a failure.
4075 PH/51 If self=fail was set on a router, the bounce message did not include the
4076 actual error message.
4082 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4084 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4085 command itself ended in error, the underlying message about the transport
4086 filter timeout was being overwritten with the pipe command error. Now the
4087 underlying error message should be appended to the second error message.
4089 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4091 PH/02 Reduce the amount of output that "make" produces by default. Full output
4092 can still be requested.
4094 PH/03 The warning log line about a condition test deferring for a "warn" verb
4095 was being output only once per connection, rather than after each
4096 occurrence (because it was using the same function as for successful
4097 "warn" verbs). This seems wrong, so I have changed it.
4099 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4100 it should not have, which might have caused a crash in the right
4101 circumstances, but probably never did.
4103 PH/04 Installed a modified version of Tony Finch's patch to make submission
4104 mode fix the return path as well as the Sender: header line, and to
4105 add a /name= option so that you can make the user's friendly name appear
4108 TF/03 Added the control = fakedefer ACL modifier.
4110 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4111 Mark Lowes for thorough testing.
4113 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4115 TK/03 Merged latest SRS patch from Miles Wilton.
4117 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4118 with the definition in sysexits.h (which is #included earlier).
4119 Fortunately, Exim does not actually use EX_OK. The code used to try to
4120 preserve the sysexits.h value, by assuming that macro definitions were
4121 scanned for macro replacements. I have been disabused of this notion,
4122 so now the code just undefines EX_OK before #including unistd.h.
4124 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4125 in the smtp transport. When a block could not be written in a single
4126 write() function, the timeout was being re-applied to each part-write.
4127 This seems wrong - if the receiver was accepting one byte at a time it
4128 would take for ever. The timeout is now adjusted when this happens. It
4129 doesn't have to be particularly precise.
4131 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4132 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4134 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4135 without a domain if the check to <postmaster@domain> fails.
4137 SC/01 Eximstats: added -xls and the ability to specify output files
4138 (patch written by Frank Heydlauf).
4140 SC/02 Eximstats: use FileHandles for outputting results.
4142 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4144 SC/04 Eximstats: fixed display of large numbers with -nvr option
4146 SC/05 Eximstats: fixed merging of reports with empty tables.
4148 SC/06 Eximstats: added the -include_original_destination flag
4150 SC/07 Eximstats: removed tabs and trailing whitespace.
4152 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4154 TK/06 MBOX spool code: Add real "From " MBOX separator line
4155 so the .eml file is really in mbox format (even though
4156 most programs do not really care). Patch from Alex Miller.
4158 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4159 The latter is generated from $received_to and is only set if the
4160 message has one envelope recipient. SA can use these headers,
4161 obviously out-of-the-box. Patch from Alex Miller.
4163 PH/08 The ${def test on a variable was returning false if the variable's
4164 value was "0", contrary to what the specification has always said!
4165 The result should be true unless the variable is empty.
4167 PH/09 The syntax error of a character other than { following "${if
4168 def:variable_name" (after optional whitespace) was not being diagnosed.
4169 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4170 accidental colon was present, for example, could give incorrect results.
4172 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4173 result is used (not including appendfile, where other changes are about
4176 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4177 This involved changing a lot of size variables from int to off_t. It
4178 should work with maildirs and everything.
4180 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4181 spamd dying while we are connected to it.
4183 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4186 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4187 function for BDB 4.3.
4189 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4191 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4192 the sources for PCRE have been re-arranged and more files are now
4195 PH/15 The code I had for printing potentially long long variables in PH/11
4196 above was not the best (it lost precision). The length of off_t variables
4197 is now inspected at build time, and an appropriate printing format (%ld
4198 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4199 to be "long long int" or "long int". This is needed for the internal
4200 formatting function string_vformat().
4202 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4203 the configuration file to be ":syslog", then the script "guesses" where
4204 the logs files are, rather than using the compiled in default. In our
4205 case the guess is not the same as the compiled default, so the script
4206 suddenly stopped working when I started to use syslog. The patch checks
4207 to see if log_file_path is "". If so, it attempts to read it from exim
4208 with no configuration file to get the compiled in version, before it
4209 falls back to the previous guessing code."
4211 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4212 implementing BATV in an Exim configuration. See NewStuff for the gory
4215 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4216 Makefile that are specific to HP-UX.
4218 PH/18 If the "use_postmaster" option was set for a recipient callout together
4219 with the "random" option, the postmaster address was used as the MAIL
4220 FROM address for the random test, but not for the subsequent recipient
4221 test. It is now used for both.
4223 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4224 patch removes a few documentation additions to RFC 3028, because the
4225 latest draft now contains them. It adds the new en;ascii-case comparator
4226 and a new error check for 8bit text in MIME parts. Comparator and
4227 require names are now matched exactly. I enabled the subaddress
4228 extension, but it is not well tested yet (read: it works for me)."
4230 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4231 rework some of the code of TK/09 above to avoid the hardwired use of
4232 "%lld" and "long long". Replaced the call to snprintf() with a call to
4235 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4236 records point to non-existent hosts", "retry timeout exceeded", and
4237 "retry time not reached for any host after a long failure period".
4239 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4240 experimental DomainKeys support:
4242 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4243 (2) On an error such as an illegally used "control", the wrong name for
4244 the control was given.
4246 These problems did NOT occur unless DomainKeys support was compiled.
4248 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4250 PH/24 Added ${if match_ip condition.
4252 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4253 databases so that it will be absolutely obvious if a crash occurs in the
4254 DB library. This is a regular occurrence (often caused by mis-matched
4257 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4258 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4259 picked up on a user's system that detects such things. There doesn't seem
4260 to be a gcc warning option for this - only an attribute that has to be
4261 put on the function's prototype. It seems that in Fedora Core 4 they have
4262 set this on a number of new functions. No doubt there will be more in due
4265 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4266 specify a transport. However, if an address that was verified by such a
4267 router was the subject of a callout, Exim crashed because it tried to
4268 read the rcpt_include_affixes from the non-existent transport. Now it
4269 just assumes that the setting of that option is false. This bug was
4270 introduced by 4.51/PH/31.
4272 PH/28 Changed -d+all to exclude +memory, because that information is very
4273 rarely of interest, but it makes the output a lot bigger. People tend to
4274 do -d+all out of habit.
4276 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4277 code in os-type was giving problems when libc.so lives in lib64, like on
4280 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4281 aren't the modern standard, and it seems that some systems' include files
4282 don't always have them. Exim was already checking for some of the newer
4283 ones like T_AAAA, and defining it itself. I've added checks for all the
4284 record types that Exim uses.
4286 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4287 not automatically generating a new one, as it is supposed to. This
4288 prevented TLS from working. If the file did exist, but contained invalid
4289 data, a new version was generated, as expected. It was only the case of a
4290 non-existent file that was broken.
4292 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4293 with a change in libdomainkeys > 0.64.
4295 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4296 from DNS. If the selector record carries the flag, it now has
4297 precedence over the domain-wide flag.
4299 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4301 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4302 the use of a port name that isn't defined in /etc/services) occurred, the
4303 message was deferred as in a normal delivery, and thus remained on the
4304 spool, instead of being failed because of the mua_wrapper setting. This
4305 is now fixed, and I tidied up some of the mua_wrapper messages at the
4308 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4309 the messages in a hash of arrays rather than using individual hashes.
4310 This is a bit cleaner and results in dramatic memory savings, albeit
4311 at a slight CPU cost.
4313 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4314 as requested by Marc Sherman.
4316 SC/10 Eximstats: added histograms for user specified patterns as requested
4319 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4321 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4322 fopen() in the content-scanning modules that did not already have it.
4328 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4329 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4331 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4333 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4335 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4336 to test Sieve filters that use "vacation".
4338 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4339 that changes the way the GnuTLS parameters are stored in the cache file.
4340 The new format can be generated externally. For backward compatibility,
4341 if the data in the cache doesn't make sense, Exim assumes it has read an
4342 old-format file, and it generates new data and writes a new file. This
4343 means that you can't go back to an older release without removing the
4346 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4347 work if there are any delivery delays because "one_time" forces the
4348 parent to be marked "delivered", so its unseen clone is never tried
4349 again. For this reason, Exim now forbids the simultaneous setting of
4352 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4353 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4354 entry if you want to know the details. The fix, however, neglected to
4355 consider the case when local delivery batching is involved. The test for
4356 "previously delivered" was not happening when checking to see if an
4357 address could be batched with a previous (undelivered) one; under
4358 certain circumstances this could lead to multiple deliveries to the same
4361 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4362 in its include files, and this causes problems building Exim.
4364 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4365 header_syntax) but Exim was just ignoring anything given after a slash.
4366 In particular, this caused confusion with an attempt to use "verify =
4367 reverse_host_lookup/defer_ok". An error is now given when options are
4368 supplied for verify items that do not have them. (Maybe reverse_host_
4369 lookup should have a defer_ok option, but that's a different point.)
4371 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4372 defined by RFC 821) to 2048, because there were problems with some AUTH
4373 commands, and RFC 1869 says the size should be increased for extended
4374 SMTP commands that take arguments.
4376 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4379 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4380 "unknown" error; now it says that the functionality isn't in the binary.
4382 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4383 an address' error message when a string expansion fails (syntax or
4384 whatever). Otherwise the password may appear in the log. Following change
4385 PH/42 below, there is no longer a chance of it appearing in a bounce
4388 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4390 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4391 of its addresses was ever tried. (Bugzilla bug #2.)
4393 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4394 the result incorrectly in the debug output. (It correctly added a newline
4395 to what was transported.)
4397 TF/01 Added $received_time.
4399 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4400 commented out examples of how to interface to a virus scanner and to
4401 SpamAssassin. Also added commented examples of av_scanner and
4402 spamd_address settings.
4404 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4405 and controls are allowed in which ACLs. There were a couple of minor
4406 errors. Some of the entries in the conditions table (which is a table of
4407 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4408 negation of where the condition IS allowed.
4410 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4412 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4413 header file does not have a version number, so I've had to invent a new
4414 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4415 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4416 radiusclient), but it was contributed by a Radius user.
4418 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4419 files or directories, for OpenSSL.
4421 PH/19 When an Exim process that is running as root has to create an Exim log
4422 file, it does so in a subprocess that runs as exim:exim so as to get the
4423 ownership right at creation (otherwise, other Exim processes might see
4424 the file with the wrong ownership). There was no test for failure of this
4425 fork() call, which would lead to the process getting stuck as it waited
4426 for a non-existent subprocess. Forks do occasionally fail when resources
4427 run out. I reviewed all the other calls to fork(); they all seem to check
4430 PH/20 When checking for unexpected SMTP input at connect time (before writing
4431 the banner), Exim was not dealing correctly with a non-positive return
4432 from the read() function. If the client had disconnected by this time,
4433 the result was a log entry for a synchronization error with an empty
4434 string after "input=" when read() returned zero. If read() returned -1
4435 (an event I could not check), uninitialized data bytes were printed.
4436 There were reports of junk text (parts of files, etc) appearing after
4439 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4441 PH/22 Added support for macro redefinition, and (re)definition in between
4442 driver and ACL definitions.
4444 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4445 forgetting to use the resulting value; it was using the unexpanded value.
4447 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4448 hadn't been configured. The fix is from Juergen Kreileder, who
4449 understands it better than I do:
4451 "Here's what I see happening with three configured cyrus_sasl
4452 authenticators configured (plain, login, cram-md5):
4454 On startup auth_cyrus_sasl_init() gets called for each of these.
4455 This means three calls to sasl_listmech() without a specified mech_list.
4456 => SASL tests which mechs of all available mechs actually work
4457 => three warnings about OTP not working
4458 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4460 With the patch, sasl_listmech() also gets called three times. But now
4461 SASL's mech_list option is set to the server_mech specified in the the
4462 authenticator. Or in other words, the answer from sasl_listmech()
4463 gets limited to just the mech you're testing for (which is different
4465 => the return list contains just 'plain' or 'login', 'cram-md5' or
4466 nothing depending on the value of ob->server_mech.
4468 I've just tested the patch: Authentication still works fine,
4469 unavailable mechs specified in the exim configuration are still
4470 caught, and the auth.log warnings about OTP are gone."
4472 PH/25 When debugging is enabled, the contents of the command line are added
4473 to the debugging output, even when log_selector=+arguments is not
4476 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4477 answer is "GNU", and only if the return is "GNU/something" is the answer
4480 PH/27 $acl_verify_message is now set immediately after the failure of a
4481 verification in an ACL, and so is available in subsequent modifiers. In
4482 particular, the message can be preserved by coding like this:
4484 warn !verify = sender
4485 set acl_m0 = $acl_verify_message
4487 Previously, $acl_verify_message was set only while expanding "message"
4488 and "log_message" when a very denied access.
4490 PH/28 Modified OS/os.c-Linux with
4492 -#ifndef OS_LOAD_AVERAGE
4493 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4495 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4496 nomenclature these days.)
4498 PH/29 Installed patch from the Sieve maintainer that adds the options
4499 sieve_useraddress and sieve_subaddress to the redirect router.
4501 PH/30 In these circumstances:
4502 . Two addresses routed to the same list of hosts;
4503 . First host does not offer TLS;
4504 . First host accepts first address;
4505 . First host gives temporary error to second address;
4506 . Second host offers TLS and a TLS session is established;
4507 . Second host accepts second address.
4508 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4509 and peerdn, if requested) that were in fact used only for the second
4512 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4513 attention to any local part prefix or suffix that was matched by the
4514 router that accepted the address. It now behaves in the same way as it
4515 does for delivery: the affixes are removed from the local part unless
4516 rcpt_include_affixes is set on the transport.
4518 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4519 timeout during the DATA phase of an incoming message.
4521 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4522 applied a patch sanctioned by the Sieve maintainer.
4524 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4525 the uid or gid is negative. A case of a negative gid caused this to be
4526 noticed. The fix allows for either to be negative.
4528 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4529 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4532 PH/36 The change PH/12 above was broken. Fixed it.
4534 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4535 the grounds that routing the same address twice would always produce the
4536 same answer. This might have been true once, but it is certainly no
4537 longer true now. Routing a child address may depend on the previous
4538 routing that produced that child. Some complicated redirection strategies
4539 went wrong when messages had multiple recipients, and made Exim's
4540 behaviour dependent on the order in which the addresses were given.
4542 I have moved the duplicate checking until after the routing is complete.
4543 Exim scans the addresses that are assigned to local and remote
4544 transports, and removes any duplicates. This means that more work will be
4545 done, as duplicates will always all be routed, but duplicates are
4546 presumably rare, so I don't expect this is of any significance.
4548 For deliveries to pipes, files, and autoreplies, the duplicate checking
4549 still happens during the routing process, since they are not going to be
4552 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4553 It corrects a timeout issue with spamd. This is Ian's comment: "The
4554 background is that sometimes spamd either never reads data from a
4555 connection it has accepted, or it never writes response data. The exiscan
4556 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4557 blindly assumes that writes won't block so it may never time out."
4559 PH/39 Allow G after quota size as well as K and M.
4561 PH/40 The value set for $authenticated_id in an authenticator may not contain
4562 binary zeroes or newlines because the value is written to log lines and
4563 to spool files. There was no check on this. Now the value is run through
4564 the string_printing() function so that such characters are converted to
4565 printable escape sequences.
4567 PH/41 $message_linecount is a new variable that contains the total number of
4568 lines in the message. Compare $body_linecount, which is the count for the
4571 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4572 bounce and delay warning messages, except in certain special cases, which
4575 (a) An SMTP error message from a remote host;
4576 (b) A message specified in a :fail: redirection;
4577 (c) A message specified in a "fail" command in a system filter;
4578 (d) A message specified in a FAIL return from the queryprogram router;
4579 (e) A message specified by the cannot_route_message router option.
4581 In these cases only, Exim does include the error details in bounce and
4582 warning messages. There are also a few cases where bland messages such
4583 as "unrouteable address" or "local delivery error" are given.
4585 PH/43 $value is now also set for the "else" part of a ${run expansion.
4587 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4588 being worked on, but at least Exim now implements the latest version to
4591 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4592 process to complete was treated as a delivery failure, a timeout while
4593 writing the message to the pipe was logged, but erroneously treated as a
4594 successful delivery. Such timeouts include transport filter timeouts. For
4595 consistency with the overall process timeout, these timeouts are now
4596 treated as errors, giving rise to delivery failures by default. However,
4597 there is now a new Boolean option for the pipe transport called
4598 timeout_defer, which, if set TRUE, converts the failures into defers for
4599 both kinds of timeout. A transport filter timeout is now identified in
4602 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4603 systems where "make" and "gmake" are different, calling "gmake" at top
4604 level broke things. I've arranged for the value of $(MAKE) to be passed
4605 from the Makefile to this script so that it can call the same version of
4609 A note about Exim versions 4.44 and 4.50
4610 ----------------------------------------
4612 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4613 changes of various kinds. As a consequence, a big documentation update was
4614 needed. This delayed the release for rather longer than seemed good, especially
4615 in the light of a couple of (minor) security issues. Therefore, the changes
4616 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4617 release. So 4.44 and 4.50 are in effect two different branches that both start
4620 I have left the 4.50 change log unchanged; it contains all the changes since
4621 4.43. The change log for 4.44 is below; many of its items are identical to
4622 those for 4.50. This seems to be the most sensible way to preserve the
4623 historical information.
4629 1. Minor wording change to the doc/README.SIEVE file.
4631 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4632 computation of the current number of files was incorrect.
4634 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4635 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4638 4. Give more explanation in the error message when the command for a transport
4639 filter fails to execute.
4641 5. There are several places where Exim runs a non-Exim command in a
4642 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4643 was being done only for the command run by the queryprogram router. It is
4644 now done for all such subprocesses. The other cases are: ${run, transport
4645 filters, and the commands run by the lmtp and pipe transports.
4647 6. Added CONFIGURE_GROUP build-time option.
4649 7. Some older OS have a limit of 256 on the maximum number of file
4650 descriptors. Exim was using setrlimit() to set 1000 as a large value
4651 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4652 systems. I've change it so that if it can't get 1000, it tries for 256.
4654 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4655 was an oversight, and furthermore, ever since the addition of extra
4656 controls (e.g. 4.43/32), the checks on when to allow different forms of
4657 "control" were broken. There should now be diagnostics for all cases when a
4658 control that does not make sense is encountered.
4660 9. Added the /retain_sender option to "control=submission".
4662 10. $recipients is now available in the predata ACL (oversight).
4664 11. Tidy the search cache before the fork to do a delivery from a message
4665 received from the command line. Otherwise the child will trigger a lookup
4666 failure and thereby defer the delivery if it tries to use (for example) a
4667 cached ldap connection that the parent has called unbind on.
4669 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4670 of $address_data from the recipient verification was clobbered by the
4671 sender verification.
4673 13. The value of address_data from a sender verification is now available in
4674 $sender_address_data in subsequent conditions in the ACL statement.
4676 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4678 15. Added a new option "connect=<time>" to callout options, to set a different
4681 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4682 was its contents. (It was OK if the option was not defined at all.)
4684 17. A "Completed" log line is now written for messages that are removed from
4685 the spool by the -Mrm option.
4687 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4688 information about exactly what failed.
4690 19. Added -dd to debug only the daemon process.
4692 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4693 handles timeouts, both on the server side and network timeouts. Renamed the
4694 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4696 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4697 It is now set to "smtps".
4699 22. $host_address is now set to the target address during the checking of
4700 ignore_target_hosts.
4702 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4703 being passed; this would have caused $sender_host_name to have been used if
4704 matching the list had actually called for a host name (not very likely,
4705 since this list is usually IP addresses). A host name is now passed as
4708 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4709 code that specifies a non-restarting handler (typically sigaction() in
4710 modern systems) in an attempt to fix a rare and obscure crash bug.
4712 25. Narrowed the window for a race in the daemon that could cause it to ignore
4713 SIGCHLD signals. This is not a major problem, because they are used only to
4714 wake it up if nothing else does.
4716 26. A malformed maildirsize file could cause Exim to calculate negative values
4717 for the mailbox size or file count. Odd effects could occur as a result.
4718 The maildirsize information is now recalculated if the size or filecount
4721 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4722 support for a long time. Removed HAVE_SYS_VFS_H.
4724 28. Installed the latest version of exipick from John Jetmore.
4726 29. In an address list, if the pattern was not a regular expression, an empty
4727 subject address (from a bounce message) matched only if the pattern was an
4728 empty string. Non-empty patterns were not even tested. This was the wrong
4729 because it is perfectly reasonable to use an empty address as part of a
4730 database query. An empty address is now tested by patterns that are
4731 lookups. However, all the other forms of pattern expect the subject to
4732 contain a local part and a domain, and therefore, for them, an empty
4733 address still always fails if the pattern is not itself empty.
4735 30. Exim went into a mad DNS loop when attempting to do a callout where the
4736 host was specified on an smtp transport, and looking it up yielded more
4737 than one IP address.
4739 31. Re-factored the code for checking spool and log partition space into a
4740 function that finds that data and another that does the check. The former
4741 is then used to implement four new variables: $spool_space, $log_space,
4742 $spool_inodes, and $log_inodes.
4744 32. The RFC2047 encoding function was originally intended for short strings
4745 such as real names; it was not keeping to the 75-character limit for
4746 encoded words that the RFC imposes. It now respects the limit, and
4747 generates multiple encoded words if necessary. To be on the safe side, I
4748 have increased the buffer size for the ${rfc2047: expansion operator from
4751 33. It is now permitted to omit both strings after an "if" condition; if the
4752 condition is true, the result is "true". As before, when the second string
4753 is omitted, a false condition yields an empty string. This makes it less
4754 cumbersome to write custom ACL and router conditions.
4756 34. Failure to deliver a bounce message always caused it to be frozen, even if
4757 there was an errors_to setting on the router. The errors_to setting is now
4760 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4761 canonical form (fully expanded) before being placed in
4762 $sender_host_address.
4764 36. The table in the code that translates DNS record types into text (T_A to
4765 "A" for instance) was missing entries for NS and CNAME. It is just possible
4766 that this could have caused confusion if both these types were looked up
4767 for the same domain, because the text type is used as part of Exim's
4768 per-process caching. But the chance of anyone hitting this buglet seems
4771 37. The dnsdb lookup has been extended in a number of ways.
4773 (1) There is a new type, "zns", which walks up the domain tree until it
4774 finds some nameserver records. It should be used with care.
4776 (2) There is a new type, "mxh", which is like "mx" except that it returns
4777 just the host names, not the priorities.
4779 (3) It is now possible to give a list of domains (or IP addresses) to be
4780 looked up. The behaviour when one of the lookups defers can be
4781 controlled by a keyword.
4783 (4) It is now possible to specify the separator character for use when
4784 multiple records are returned.
4786 38. The dnslists ACL condition has been extended: it is now possible to supply
4787 a list of IP addresses and/or domains to be looked up in a particular DNS
4790 39. Added log_selector=+queue_time_overall.
4792 40. When running the queue in the test harness, wait just a tad after forking a
4793 delivery process, to get repeatability of debugging output.
4795 41. Include certificate and key file names in error message when GnuTLS fails
4796 to set them up, because the GnuTLS error message doesn't include the name
4797 of the failing file when there is a problem reading it.
4799 42. Allow both -bf and -bF in the same test run.
4801 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4803 44. The "Exiscan patch" is now merged into the mainline Exim source.
4805 45. Sometimes the final signoff response after QUIT could fail to get
4806 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4807 < 0 before doing a fflush(). This bug looks as though it goes back to the
4808 introduction of TLS in release 3.20, but "sometimes" must have been rare
4809 because the tests only now provoked it.
4811 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4812 (this can affect the format of dates).
4814 47. exim_tidydb, when checking for the continued existence of a message for
4815 which it has found a message-specific retry record, was not finding
4816 messages that were in split spool directories. Consequently, it was
4817 deleting retry records that should have stayed in existence.
4819 48. Steve fixed some bugs in eximstats.
4821 49. The SPA authentication driver was not abandoning authentication and moving
4822 on to the next authenticator when an expansion was forced to fail,
4823 contradicting the general specification for all authenticators. Instead it
4824 was generating a temporary error. It now behaves as specified.
4826 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4827 (the order specifies the preference for clients). The order is now AES256,
4828 AES128, 3DES, ARCFOUR128.
4830 51. Small patch to Sieve code - explicitly set From: when generating an
4833 52. Exim crashed if a remote delivery caused a very long error message to be
4834 recorded - for instance if somebody sent an entire SpamAssassin report back
4835 as a large number of 550 error lines. This bug was coincidentally fixed by
4836 increasing the size of one of Exim's internal buffers (big_buffer) that
4837 happened as part of the Exiscan merge. However, to be on the safe side, I
4838 have made the code more robust (and fixed the comments that describe what
4841 53. Now that there can be additional text after "Completed" in log lines (if
4842 the queue_time_overall log selector is set), a one-byte patch to exigrep
4843 was needed to allow it to recognize "Completed" as not the last thing in
4846 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4847 patch that reportedly fixes this has been added. I am not expert enough to
4848 create a test for it. This is what the patch creator wrote:
4850 "I found a little strange behaviour of ldap code when working with
4851 Windows 2003 AD Domain, where users was placed in more than one
4852 Organization Units. When I tried to give exim partial DN, the exit code
4853 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4854 But simultaneously result of request was absolutely normal ldap result,
4855 so I produce this patch..."
4857 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4858 REFERENCE, so I have modified the code to exclude the patch when that macro
4861 55. Some experimental protocols are using DNS PTR records for new purposes. The
4862 keys for these records are domain names, not reversed IP addresses. The
4863 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4864 leaves it alone. Component reversal etc. now happens only for IP addresses.
4867 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4869 57. Double the size of the debug message buffer (to 2048) so that more of very
4870 long debug lines gets shown.
4872 58. The exicyclog utility now does better if the number of log files to keep
4873 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4875 59. Two changes related to the smtp_active_hostname option:
4877 (1) $smtp_active_hostname is now available as a variable.
4878 (2) The default for smtp_banner uses $smtp_active_hostname instead
4879 of $primary_hostname.
4881 60. The host_aton() function is supposed to be passed a string that is known
4882 to be a valid IP address. However, in the case of IPv6 addresses, it was
4883 not checking this. This is a hostage to fortune. Exim now panics and dies
4884 if the condition is not met. A case was found where this could be provoked
4885 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4886 components; fortuitously, this particular loophole had already been fixed
4887 by change 4.50/55 above.
4889 If there are any other similar loopholes, the new check in host_aton()
4890 itself should stop them being exploited. The report I received stated that
4891 data on the command line could provoke the exploit when Exim was running as
4892 exim, but did not say which command line option was involved. All I could
4893 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4894 running as the user.
4897 61. There was a buffer overflow vulnerability in the SPA authentication code
4898 (which came originally from the Samba project). I have added a test to the
4899 spa_base64_to_bits() function which I hope fixes it.
4902 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4903 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4905 63. The daemon start-up calls getloadavg() while still root for those OS that
4906 need the first call to be done as root, but it missed one case: when
4907 deliver_queue_load_max is set with deliver_drop_privilege. This is
4908 necessary for the benefit of the queue runner, because there is no re-exec
4909 when deliver_drop_privilege is set.
4911 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4912 This has been fixed.
4914 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4915 was in use, was not putting the data itself into the right store pool;
4916 consequently, it could be overwritten for a subsequent message in the same
4917 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4920 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4922 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4923 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4924 the function were treating the return as a boolean value, which happened to
4925 work because 0=false and not-0=true, but is not correct code.
4927 68. The host_aton() function was not handling scoped IPv6 addresses (those
4928 with, for example, "%eth0" on the end) correctly.
4930 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4931 negated items (that is, ~something) in unsigned ints. Some compilers
4932 apparently mutter when there is no cast.
4934 70. If an address verification called from an ACL failed, and did not produce a
4935 user-specific message (i.e. there was only a "system" message), nothing was
4936 put in $acl_verify_message. In this situation, it now puts the system
4939 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4940 session; change 4.31/43 added the unwanted input to the log line - except
4941 that it did not do this in the start of session case. It now does.
4943 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4944 This is wrong and can cause the other end to generate a synchronization
4945 error if it is another Exim or anything else that does the synchronization
4946 check. A QUIT command is no longer sent after a timeout.
4948 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4949 during host lookups.
4951 74. The defer_ok option of callout verification was not working if it was used
4952 when verifying addresses in header lines, that is, for this case:
4954 verify = header_sender/callout=defer_ok
4956 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4957 those file descriptors could be used for SMTP connections. If anything
4958 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4959 could be sent to the SMTP client, causing chaos. The daemon now opens
4960 stdin, stdout, and stderr to /dev/null when it puts itself into the
4963 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4964 log by default. The user can override this with suitable Perl magic.
4966 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4967 the log message when discard triggers, was not working for the DATA ACL or
4968 for the non-SMTP ACL.
4970 78. Error message wording change in sieve.c.
4972 79. If smtp_accept_max_per_host was set, the number of connections could be
4973 restricted to fewer than expected, because the daemon was trying to set up
4974 a new connection before checking whether the processes handling previous
4975 connections had finished. The check for completed processes is now done
4976 earlier. On busy systems, this bug wouldn't be noticed because something
4977 else would have woken the daemon, and it would have reaped the completed
4980 80. If a message was submitted locally by a user whose login name contained one
4981 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4982 It caused a spool format error. I have fixed the spool reading code. A
4983 related problem was that the "from" clause in the Received: line became
4984 illegal because of the space(s). It is now covered by ${quote_local_part.
4986 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4989 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4991 83. Patch from Sieve maintainer to fix unterminated string problem in
4992 "vacation" handling.
4994 84. Some minor changes to the Linux configuration files to help with other
4995 OS variants using glibc.
4997 85. One more patch for Sieve to update vacation handling to latest spec.
5000 ----------------------------------------------------
5001 See the note above about the 4.44 and 4.50 releases.
5002 ----------------------------------------------------
5008 1. Change 4.43/35 introduced a bug that caused file counts to be
5009 incorrectly computed when quota_filecount was set in an appendfile
5012 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5013 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5016 3. Give more explanation in the error message when the command for a transport
5017 filter fails to execute.
5019 4. There are several places where Exim runs a non-Exim command in a
5020 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5021 was being done only for the command run by the queryprogram router. It is
5022 now done for all such subprocesses. The other cases are: ${run, transport
5023 filters, and the commands run by the lmtp and pipe transports.
5025 5. Some older OS have a limit of 256 on the maximum number of file
5026 descriptors. Exim was using setrlimit() to set 1000 as a large value
5027 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5028 systems. I've change it so that if it can't get 1000, it tries for 256.
5030 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5031 was an oversight, and furthermore, ever since the addition of extra
5032 controls (e.g. 4.43/32), the checks on when to allow different forms of
5033 "control" were broken. There should now be diagnostics for all cases when a
5034 control that does not make sense is encountered.
5036 7. $recipients is now available in the predata ACL (oversight).
5038 8. Tidy the search cache before the fork to do a delivery from a message
5039 received from the command line. Otherwise the child will trigger a lookup
5040 failure and thereby defer the delivery if it tries to use (for example) a
5041 cached ldap connection that the parent has called unbind on.
5043 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5044 of $address_data from the recipient verification was clobbered by the
5045 sender verification.
5047 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5048 was its contents. (It was OK if the option was not defined at all.)
5050 11. A "Completed" log line is now written for messages that are removed from
5051 the spool by the -Mrm option.
5053 12. $host_address is now set to the target address during the checking of
5054 ignore_target_hosts.
5056 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5057 being passed; this would have caused $sender_host_name to have been used if
5058 matching the list had actually called for a host name (not very likely,
5059 since this list is usually IP addresses). A host name is now passed as
5062 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5063 code that specifies a non-restarting handler (typically sigaction() in
5064 modern systems) in an attempt to fix a rare and obscure crash bug.
5066 15. Narrowed the window for a race in the daemon that could cause it to ignore
5067 SIGCHLD signals. This is not a major problem, because they are used only to
5068 wake it up if nothing else does.
5070 16. A malformed maildirsize file could cause Exim to calculate negative values
5071 for the mailbox size or file count. Odd effects could occur as a result.
5072 The maildirsize information is now recalculated if the size or filecount
5075 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5076 support for a long time. Removed HAVE_SYS_VFS_H.
5078 18. Updated exipick to current release from John Jetmore.
5080 19. Allow an empty sender to be matched against a lookup in an address list.
5081 Previously the only cases considered were a regular expression, or an
5084 20. Exim went into a mad DNS lookup loop when doing a callout where the
5085 host was specified on the transport, if the DNS lookup yielded more than
5088 21. The RFC2047 encoding function was originally intended for short strings
5089 such as real names; it was not keeping to the 75-character limit for
5090 encoded words that the RFC imposes. It now respects the limit, and
5091 generates multiple encoded words if necessary. To be on the safe side, I
5092 have increased the buffer size for the ${rfc2047: expansion operator from
5095 22. Failure to deliver a bounce message always caused it to be frozen, even if
5096 there was an errors_to setting on the router. The errors_to setting is now
5099 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5100 canonical form (fully expanded) before being placed in
5101 $sender_host_address.
5103 24. Updated eximstats to version 1.33
5105 25. Include certificate and key file names in error message when GnuTLS fails
5106 to set them up, because the GnuTLS error message doesn't include the name
5107 of the failing file when there is a problem reading it.
5109 26. Expand error message when OpenSSL has problems setting up cert/key files.
5112 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5113 (this can affect the format of dates).
5115 28. exim_tidydb, when checking for the continued existence of a message for
5116 which it has found a message-specific retry record, was not finding
5117 messages that were in split spool directories. Consequently, it was
5118 deleting retry records that should have stayed in existence.
5120 29. eximstats updated to version 1.35
5121 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5122 1.35 - bugfix such that pie charts by volume are generated correctly
5124 30. The SPA authentication driver was not abandoning authentication and moving
5125 on to the next authenticator when an expansion was forced to fail,
5126 contradicting the general specification for all authenticators. Instead it
5127 was generating a temporary error. It now behaves as specified.
5129 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5130 (the order specifies the preference for clients). The order is now AES256,
5131 AES128, 3DES, ARCFOUR128.
5133 31. Small patch to Sieve code - explicitly set From: when generating an
5136 32. Exim crashed if a remote delivery caused a very long error message to be
5137 recorded - for instance if somebody sent an entire SpamAssassin report back
5138 as a large number of 550 error lines. This bug was coincidentally fixed by
5139 increasing the size of one of Exim's internal buffers (big_buffer) that
5140 happened as part of the Exiscan merge. However, to be on the safe side, I
5141 have made the code more robust (and fixed the comments that describe what
5144 33. Some experimental protocols are using DNS PTR records for new purposes. The
5145 keys for these records are domain names, not reversed IP addresses. The
5146 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5147 leaves it alone. Component reversal etc. now happens only for IP addresses.
5150 34. The host_aton() function is supposed to be passed a string that is known
5151 to be a valid IP address. However, in the case of IPv6 addresses, it was
5152 not checking this. This is a hostage to fortune. Exim now panics and dies
5153 if the condition is not met. A case was found where this could be provoked
5154 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5155 components; fortuitously, this particular loophole had already been fixed
5156 by change 4.50/55 or 4.44/33 above.
5158 If there are any other similar loopholes, the new check in host_aton()
5159 itself should stop them being exploited. The report I received stated that
5160 data on the command line could provoke the exploit when Exim was running as
5161 exim, but did not say which command line option was involved. All I could
5162 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5163 running as the user.
5166 35. There was a buffer overflow vulnerability in the SPA authentication code
5167 (which came originally from the Samba project). I have added a test to the
5168 spa_base64_to_bits() function which I hope fixes it.
5171 36. The daemon start-up calls getloadavg() while still root for those OS that
5172 need the first call to be done as root, but it missed one case: when
5173 deliver_queue_load_max is set with deliver_drop_privilege. This is
5174 necessary for the benefit of the queue runner, because there is no re-exec
5175 when deliver_drop_privilege is set.
5177 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5178 was in use, was not putting the data itself into the right store pool;
5179 consequently, it could be overwritten for a subsequent message in the same
5180 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5183 38. Sometimes the final signoff response after QUIT could fail to get
5184 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5185 < 0 before doing a fflush(). This bug looks as though it goes back to the
5186 introduction of TLS in release 3.20, but "sometimes" must have been rare
5187 because the tests only now provoked it.
5193 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5194 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5195 Now it returns an integer. A number of calls were still expecting a T/F
5196 return. Fortuitously, in all cases, the tests worked in OK situations,
5197 which is the norm. However, things would have gone wrong on any write
5198 failures on the smtp file descriptor. This function is used when sending
5199 messages over SMTP and also when doing verify callouts.
5201 2. When Exim is called to do synchronous delivery of a locally submitted
5202 message (the -odf or -odi options), it no longer closes stderr before doing
5205 3. Implemented the mua_wrapper option.
5207 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5209 5. Implemented the functions header_remove(), header_testname(),
5210 header_add_at_position(), and receive_remove_recipient(), and exported them
5213 6. If an ACL "warn" statement specified the addition of headers, Exim already
5214 inserted X-ACL-Warn: at the start if there was no header name. However, it
5215 was not making this test for the second and subsequent header lines if
5216 there were newlines in the string. This meant that an invalid header could
5217 be inserted if Exim was badly configured.
5219 7. Allow an ACL "warn" statement to add header lines at the start or after all
5220 the Received: headers, as well as at the end.
5222 8. Added the rcpt_4xx retry error code.
5224 9. Added postmaster_mailfrom=xxx to callout verification option.
5226 10. Added mailfrom=xxxx to the callout verification option, for verify=
5229 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5230 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5232 12. Inserted some casts to stop certain compilers warning when using pointer
5233 differences as field lengths or precisions in printf-type calls (mostly
5234 affecting debugging statements).
5236 13. Added optional readline() support for -be (dynamically loaded).
5238 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5239 same clock tick as a message's arrival, so that its received time was the
5240 same as the "first fail" time on the retry record, and that message
5241 remained on the queue past the ultimate address timeout, every queue runner
5242 would try a delivery (because it was past the ultimate address timeout) but
5243 after another failure, the ultimate address timeout, which should have then
5244 bounced the address, did not kick in. This was a "< instead of <=" error;
5245 in most cases the first failure would have been in the next clock tick
5246 after the received time, and all would be well.
5248 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5249 being recognized when the domain list was tested by the match_domain
5250 condition in an expansion string.
5252 16. Added the ${str2b64: operator.
5254 17. Exim was always calling setrlimit() to set a large limit for the number of
5255 processes, without checking whether the existing limit was already
5256 adequate. (It did check for the limit on file descriptors.) Furthermore,
5257 errors from getrlimit() and setrlimit() were being ignored. Now they are
5258 logged to the main and panic logs, but Exim does carry on, to try to do its
5259 job under whatever limits there are.
5261 18. Imported PCRE 5.0.
5263 19. Trivial typo in log message " temporarily refused connection" (the leading
5266 20. If the log selector return_path_on_delivery was set and an address was
5267 redirected to /dev/null, the delivery process crashed because it assumed
5268 that a return path would always be set for a "successful" delivery. In this
5269 case, the whole delivery is bypassed as an optimization, and therefore no
5272 21. Internal re-arrangement: the function for sending a challenge and reading
5273 a response while authentication was assuming a zero-terminated challenge
5274 string. It's now changed to take a pointer and a length, to allow for
5275 binary data in such strings.
5277 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5279 23. Exim was not respecting finduser_retries when seeking the login of the
5280 uid under which it was called; it was always trying 10 times. (The default
5281 setting of finduser_retries is zero.) Also, it was sleeping after the final
5282 failure, which is pointless.
5284 24. Implemented tls_on_connect_ports.
5286 25. Implemented acl_smtp_predata.
5288 26. If the domain in control=submission is set empty, Exim assumes that the
5289 authenticated id is a complete email address when it generates From: or
5290 Sender: header lines.
5292 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5293 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5294 chown and chgrp in /bin and hostname in /usr/bin.
5296 28. Exim was keeping the "process log" file open after each use, just as it
5297 does for the main log. This opens the possibility of it remaining open for
5298 long periods when the USR1 signal hits a daemon. Occasional processlog
5299 errors were reported, that could have been caused by this. Anyway, it seems
5300 much more sensible not to leave this file open at all, so that is what now
5303 29. The long-running daemon process does not normally write to the log once it
5304 has entered its main loop, and it closes the log before doing so. This is
5305 so that log files can straightforwardly be renamed and moved. However,
5306 there are a couple of unusual error situations where the daemon does write
5307 log entries, and I had neglected to close the log afterwards.
5309 30. The text of an SMTP error response that was received during a remote
5310 delivery was being truncated at 512 bytes. This is too short for some of
5311 the long messages that one sometimes sees. I've increased the limit to
5314 31. It is now possible to make retry rules that apply only when a message has a
5315 specific sender, in particular, an empty sender.
5317 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5318 it possible to be selective about when SMTP synchronization is enforced.
5320 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5322 32. Implemented hosts_connection_nolog.
5324 33. Added an ACL for QUIT.
5326 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5329 35. Added mailbox_size and mailbox_filecount to appendfile.
5331 36. Added control = no_multiline_responses to ACLs.
5333 37. There was a bug in the logic of the code that waits for the clock to tick
5334 in the case where the clock went backwards by a substantial amount such
5335 that the microsecond fraction of "now" was more than the microsecond
5336 fraction of "then" (but the whole seconds number was less).
5338 38. Added support for the libradius Radius client library this is found on
5339 FreeBSD (previously only the radiusclient library was supported).
5345 1. When certain lookups returned multiple values in the form name=value, the
5346 quoting of the values was not always being done properly. Specifically:
5347 (a) If the value started with a double quote, but contained no whitespace,
5349 (b) If the value contained whitespace other than a space character (i.e.
5350 tabs or newlines or carriage returns) it was not quoted.
5351 This fix has been applied to the mysql and pgsql lookups by writing a
5352 separate quoting function and calling it from the lookup code. The fix
5353 should probably also be applied to nisplus, ibase and oracle lookups, but
5354 since I cannot test any of those, I have not disturbed their existing code.
5356 2. A hit in the callout cache for a specific address caused a log line with no
5357 reason for rejecting RCPT. Now it says "Previous (cached) callout
5358 verification failure".
5360 3. There was an off-by-one bug in the queryprogram router. An over-long
5361 return line was truncated at 256 instead of 255 characters, thereby
5362 overflowing its buffer with the terminating zero. As well as fixing this, I
5363 have increased the buffer size to 1024 (and made a note to document this).
5365 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5366 when Exim is waiting for an SMTP response from a remote server, Exim
5367 restarts its select() call on the socket, thereby resetting its timeout.
5368 This is not a problem when such interrupts are rare. Somebody set up a cron
5369 job to run exiwhat every 2 minutes, which is less than the normal select()
5370 timeout (5 or 10 minutes). This meant that the select() timeout never
5371 kicked in because it was always reset. I have fixed this by comparing the
5372 time when an interrupt arrives with the time at the start of the first call
5373 to select(). If more time than the timeout has elapsed, the interrupt is
5374 treated as a timeout.
5376 5. Some internal re-factoring in preparation for the addition of Sieve
5377 extensions (by MH). In particular, the "personal" test is moved to a
5378 separate function, and given an option for scanning Cc: and Bcc: (which is
5379 not set for Exim filters).
5381 6. When Exim created an email address using the login of the caller as the
5382 local part (e.g. when creating a From: or Sender: header line), it was not
5383 quoting the local part when it contained special characters such as @.
5385 7. Installed new OpenBSD configuration files.
5387 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5388 try to make them clearer.
5390 9. Callout options, other than the timeout value, were being ignored when
5391 verifying sender addresses in header lines. For example, when using
5393 verify = header_sender/callout=no_cache
5395 the cache was (incorrectly) being used.
5397 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5398 only the Cygwin environment.
5400 11. When return_path_on_delivery was set as a log selector, if different remote
5401 addresses in the same message used different return paths and parallel
5402 remote delivery occurred, the wrong values would sometimes be logged.
5403 (Whenever a remote delivery process finished, the return path value from
5404 the most recently started remote delivery process was logged.)
5406 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5407 lines when AUTH and/or TLS are in use. This is the "received protocol"
5408 field. Exim used to use "asmtp" for authenticated SMTP, without any
5409 indication (in the protocol name) for TLS use. Now it follows the RFC and
5410 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5411 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5412 names appear in log lines as well as in Received: header lines.
5414 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5415 extensions, and comparison tests, and to fix some bugs.
5417 14. Changes to the "personal" filter test:
5419 (1) The test was buggy in that it was just doing the equivalent of
5420 "contains" tests on header lines. For example, if a user's address was
5421 anne@some.where, the "personal" test would incorrectly be true for
5423 To: susanne@some.where
5425 This test is now done by extracting each address from the header in turn,
5426 and checking the entire address. Other tests that are part of "personal"
5427 are now done using regular expressions (for example, to check local parts
5428 of addresses in From: header lines).
5430 (2) The list of non-personal local parts in From: addresses has been
5431 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5432 taken from the Sieve specification recommendations.
5434 (3) If the message contains any header line starting with "List-" it is
5435 treated as non-personal.
5437 (4) The test for "circular" in the Subject: header line has been removed
5438 because it now seems ill-conceived.
5440 15. Minor typos in src/EDITME comments corrected.
5442 16. Installed latest exipick from John Jetmore.
5444 17. If headers_add on a router specified a text string that was too long for
5445 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5446 of string_sprintf() is now avoided.
5448 18. $message_body_size was not set (it was always zero) when running the DATA
5449 ACL and the local_scan() function.
5451 19. For the "mail" command in an Exim filter, no default was being set for
5452 the once_repeat time, causing a random time value to be used if "once" was
5453 specified. (If the value happened to be <= 0, no repeat happened.) The
5454 default is now 0s, meaning "never repeat". The "vacation" command was OK
5455 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5456 (I found it when inspecting the code).
5458 20. There is now an overall timeout for performing a callout verification. It
5459 defaults to 4 times the callout timeout, which applies to individual SMTP
5460 commands during the callout. The overall timeout applies when there is more
5461 than one host that can be tried. The timeout is checked before trying the
5462 next host. This prevents very long delays if there are a large number of
5463 hosts and all are timing out (e.g. when the network connections are timing
5464 out). The value of the overall timeout can be changed by specifying an
5465 additional sub-option for "callout", called "maxwait". For example:
5467 verify = sender/callout=5s,maxwait=20s
5469 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5470 to the end before writing, but this should make it even safer).
5472 22. Exim was forgetting that it had advertised PIPELINING for the second and
5473 subsequent messages on an SMTP connection. It was also not resetting its
5474 memory on STARTTLS and an internal HELO.
5476 23. When Exim logs an SMTP synchronization error within a session, it now
5477 records whether PIPELINING has been advertised or not.
5479 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5480 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5481 rather than long int.
5483 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5485 26. Added the never_mail option to autoreply.
5491 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5492 crash if the getsockname() call failed; this can happen if a connection is
5493 closed very soon after it is established. The problem was simply in the
5494 order in which certain operations were done, causing Exim to try to write
5495 to the SMTP stream before it had set up the file descriptor. The bug has
5496 been fixed by making things happen in the correct order.
5502 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5503 before the connection was closed, thus losing the rejection response.
5505 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5506 some early Solaris releases, but causes trouble in current releases where
5507 socklen_t is defined.
5509 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5512 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5515 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5516 that releasing the top of it at the end releases what was used for sub-
5517 expansions (unless the block got too big). However, discard this block if
5518 the first thing is a variable or header, so that we can use its block when
5519 it is dynamic (useful for very large $message_headers, for example).
5521 6. Lookups now cache *every* query, not just the most recent. A new, separate
5522 store pool is used for this. It can be recovered when all lookup caches are
5523 flushed. Lookups now release memory at the end of their result strings.
5524 This has involved some general refactoring of the lookup sources.
5526 7. Some code has been added to the store_xxx() functions to reduce the amount
5527 of flapping under certain conditions.
5529 8. log_incoming_interface used to affect only the <= reception log lines. Now
5530 it causes the local interface and port to be added to several more SMTP log
5531 lines, for example "SMTP connection from", and rejection lines.
5533 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5535 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5537 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5538 could be overwritten at the end of the current message (or the start of a
5539 new message if it was set in a HELO ACL). The value is now preserved for
5540 the duration of the SMTP connection.
5542 12. If a transport had a headers_rewrite setting, and a matching header line
5543 contained an unqualified address, that address was qualified, even if it
5544 did not match any rewriting rules. The underlying bug was that the values
5545 of the flags that permit the existence of unqualified sender and recipient
5546 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5547 non-local messages, and by -bnq for local messages) were not being
5548 preserved with the message after it was received.
5550 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5551 "next input=" as part of the text comprising the host identity instead of
5552 the correct text. The code was using the same buffer for two different
5553 strings. However, depending on which order the printing function evaluated
5554 its arguments, the bug did not always show up. Under Linux, for example, my
5555 test suite worked just fine.
5557 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5558 doesn't work with some older versions of Perl. It has been changed to "my",
5559 which in any case is probably the better facility to use.
5561 15. A really picky compiler found some instances of statements for creating
5562 error messages that either had too many or two few arguments for the format
5565 16. The size of the buffer for calls to the DNS resolver has been increased
5566 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5567 for addresses that have a lot of PTR records. This alleviates a problem; it
5568 does not fully solve it.
5570 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5571 buffer now truncates the list and logs the incident, which is the same
5572 action as happens when Exim is looking up a host name and its aliases.
5573 Previously in this situation something unpredictable would happen;
5574 sometimes it was "internal error: store_reset failed".
5576 18. If a server dropped the connection unexpectedly when an Exim client was
5577 using GnuTLS and trying to read a response, the client delivery process
5578 crashed while trying to generate an error log message.
5580 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5581 string, for example:
5583 warn message = H1: something\nH2: something
5585 the text was added as a single header line from Exim's point of view
5586 though it ended up OK in the delivered message. However, searching for the
5587 second and subsequent header lines using $h_h2: did not work. This has been
5588 fixed. Similarly, if a system filter added multiple headers in this way,
5589 the routers could not see them.
5591 20. Expanded the error message when iplsearch is called with an invalid key to
5592 suggest using net-iplsearch in a host list.
5594 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5595 ACLs are no longer actually imposed (and a message to that effect is
5598 22. If a "gecos" field in a passwd entry contained escaped characters, in
5599 particular, if it contained a \" sequence, Exim got it wrong when building
5600 a From: or a Sender: header from that name. A second bug also caused
5601 incorrect handling when an unquoted " was present following a character
5602 that needed quoting.
5604 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5605 was not being matched caselessly.
5607 24. Arranged for all hyphens in the exim.8 source to be escaped with
5610 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5611 properly. Recipient callout cache records were still being keyed to include
5612 the sender, even when use_sender was set false. This led to far more
5613 callouts that were necessary. The sender is no longer included in the key
5614 when use_sender is false.
5616 26. Added "control = submission" modifier to ACLs.
5618 27. Added the ${base62d: operator to decode base 62 numbers.
5620 28. dnsdb lookups can now access SRV records.
5622 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5623 the configuration file.
5625 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5626 (-v) mode. This makes the output for a verbose queue run more intelligible.
5628 31. Added a use_postmaster feature to recipient callouts.
5630 32. Added the $body_zerocount variable, containing the number of binary zero
5631 bytes in the message body.
5633 33. The time of last modification of the "new" subdirectory is now used as the
5634 "mailbox time last read" when there is a quota error for a maildir
5637 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5639 35. Added +ignore_unknown as a special item in host lists.
5641 36. Code for decoding IPv6 addresses in host lists is now included, even if
5642 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5643 address was recognized as an IP address, but was then not correctly decoded
5644 into binary, causing unexpected and incorrect effects when compared with
5651 1. Very minor rewording of debugging text in manualroute to say "list of
5652 hosts" instead of "hostlist".
5654 2. If verify=header_syntax was set, and a header line with an unqualified
5655 address (no domain) and a large number of spaces between the end of the
5656 name and the colon was received, the reception process suffered a buffer
5657 overflow, and (when I tested it) crashed. This was caused by some obsolete
5658 code that should have been removed. The fix is to remove it!
5660 3. When running in the test harness, delay a bit after writing a bounce
5661 message to get a bit more predictability in the log output.
5663 4. Added a call to search_tidyup() just before forking a reception process. In
5664 theory, someone could use a lookup in the expansion of smtp_accept_max_
5665 per_host which, without the tidyup, could leave open a database connection.
5667 5. Added the variables $recipient_data and $sender_data which get set from a
5668 lookup success in an ACL "recipients" or "senders" condition, or a router
5669 "senders" option, similar to $domain_data and $local_part_data.
5671 6. Moved the writing of debug_print from before to after the "senders" test
5674 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5675 problems for message scanning, either using a data ACL, or using
5676 local_scan() because the Received: header was not generated till after they
5677 were called (in order to set the time as the time of reception completion).
5678 I have revised the way this works. The header is now generated after the
5679 body is received, but before the ACL or local_scan() are called. After they
5680 are run, the timestamp in the header is updated.
5686 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5687 before starting a queue runner without re-exec. This happened only when
5688 deliver_drop_privilege was set or when the Exim user was set to root. The
5689 effect of the bug was that timeouts during subsequent deliveries caused
5690 crashes instead of being properly handled. The handler is now left at its
5691 default (and expected) setting.
5693 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5694 message, again when deliver_drop_privilege is set or Exim is run as root.
5695 The bug described in (1) was not present in this case, but the tidying up
5696 of the other signals was missing. I have made the two cases consistent.
5698 3. The ignore_target_hosts setting on a manualroute router was being ignored
5699 for hosts that were looked up using the /MX notation.
5701 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5704 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5705 operated on the sender address. After changing the $sender_address to <>
5706 for the sender address verify, Exim was re-instated it as the original
5707 (before rewriting) address, but remembering that it had rewritten it, so it
5708 wasn't rewriting it again. This bug also had the effect of breaking the
5709 sender address verification caching when the sender address was rewritten.
5711 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5712 This has been changed so that if the ip literal address matches
5713 ignore_target_hosts, the router declines.
5715 7. Added expansion conditions match_domain, match_address, and match_local_
5716 part (NOT match_host).
5718 8. The placeholder for the Received: header didn't have a length field set.
5720 9. Added code to Exim itself and to exim_lock to test for a specific race
5721 condition that could lead to file corruption when using MBX delivery. The
5722 issue is with the lockfile that is created in /tmp. If this file is removed
5723 after a process has opened it but before that process has acquired a lock,
5724 there is the potential for a second process to recreate the file and also
5725 acquire a lock. This could lead to two Exim processes writing to the file
5726 at the same time. The added code performs the same test as UW imapd; it
5727 checks after acquiring the lock that its file descriptor still refers to
5728 the same named file.
5730 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5731 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5734 11. Added the smtp_active_hostname option. If used, this will typically be made
5735 to depend on the incoming interface address. Because $interface_address is
5736 not set up until the daemon has forked a reception process, error responses
5737 that can happen earlier (such as "too many connections") no longer contain
5740 12. If an expansion in a condition on a "warn" statement fails because a lookup
5741 defers, the "warn" statement is abandoned, and the next ACL statement is
5742 processed. Previously this caused the whole ACL to be aborted.
5744 13. Added the iplsearch lookup type.
5746 14. Added ident_timeout as a log selector.
5748 15. Added tls_certificate_verified as a log selector.
5750 16. Added a global option tls_require_ciphers (compare the smtp transport
5751 option of the same name). This controls incoming TLS connections.
5753 17. I finally figured out how to make tls_require_ciphers do a similar thing
5754 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5755 before starting the TLS session.
5757 18. Tabs are now shown as \t in -bP output.
5759 19. If the log selector return_path_on_delivery was set, Exim crashed when
5760 bouncing a message because it had too many Received: header lines.
5762 20. If two routers both had headers_remove settings, and the first one included
5763 a superfluous trailing colon, the final name in the first list and the
5764 first name in the second list were incorrectly joined into one item (with a
5765 colon in the middle).
5771 1. Added -C and -D options to the exinext utility, mainly to make it easier
5772 to include in the automated testing, but these could be helpful when
5773 multiple configurations are in use.
5775 2. The exinext utility was not formatting the output nicely when there was
5776 an alternate port involved in the retry record key, nor when there was a
5777 message id as well (for retries that were specific to a specific message
5778 and a specific host). It was also confused by IPv6 addresses, because of
5779 the additional colons they contain. I have fixed the IPv4 problem, and
5780 patched it up to do a reasonable job for IPv6.
5782 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5783 delivery, the log line now contains "pipelined" if PIPELINING was used.
5785 4. An SMTP transport process used to panic and die if the bind() call to set
5786 an explicit outgoing interface failed. This has been changed; it is now
5787 treated in the same way as a connect() failure.
5789 5. A reference to $sender_host_name in the part of a conditional expansion
5790 that was being skipped was still causing a DNS lookup. This no longer
5793 6. The def: expansion condition was not recognizing references to header lines
5794 that used bh_ and bheader_.
5796 7. Added the _cache feature to named lists.
5798 8. The code for checking quota_filecount in the appendfile transport was
5799 allowing one more file than it should have been.
5801 9. For compatibility with Sendmail, the command line option
5809 and sets the incoming protocol and host name (for trusted callers). The
5810 host name and its colon can be omitted when only the protocol is to be set.
5811 Note the Exim already has two private options, -pd and -ps, that refer to
5812 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5813 "s", but I don't think that's a major issue.
5815 10. A number of refactoring changes to the code, none of which should affect
5818 (a) The number of logging options was getting close to filling up the
5819 32-bit word that was used as a bit map. I have split them into two classes:
5820 those that are passed in the argument to log_write(), and those that are
5821 only ever tested independently outside of that function. These are now in
5822 separate 32-bit words, so there is plenty of room for expansion again.
5823 There is no change in the user interface or the logging behaviour.
5825 (b) When building, for example, log lines, the code previously used a
5826 macro that called string_cat() twice, in order to add two strings. This is
5827 not really sufficiently general. Furthermore, there was one instance where
5828 it was actually wrong because one of the argument was used twice, and in
5829 one call a function was used. (As it happened, calling the function twice
5830 did not affect the overall behaviour.) The macro has been replaced by a
5831 function that can join an arbitrary number of extra strings onto a growing
5834 (c) The code for expansion conditions now uses a table and a binary chop
5835 instead of a serial search (which was left over from when there were very
5836 few conditions). Also, it now recognizes conditions like "pam" even when
5837 the relevant support is not compiled in: a suitably worded error message is
5838 given if an attempt is made to use such a condition.
5840 11. Added ${time_interval:xxxxx}.
5842 12. A bug was causing one of the ddress fields not to be passed back correctly
5843 from remote delivery subprocesses. The field in question was not being
5844 subsequently used, so this caused to problems in practice.
5846 13. Added new log selectors queue_time and deliver_time.
5848 14. Might have fixed a bug in maildirsizefile handling that threw up
5849 "unexpected character" debug warnings, and recalculated the data
5850 unnecessarily. In any case, I expanded the warning message to give more
5853 15. Added the message "Restricted characters in address" to the statements in
5854 the default ACL that block characters like @ and % in local parts.
5856 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5857 Three changes have been made:
5859 (a) There was a serious bug; a negative response to MAIL caused the whole
5860 recipient domain to be cached as invalid, thereby blocking all messages
5861 to all local parts at the same domain, from all senders. This bug has
5862 been fixed. The domain is no longer cached after a negative response to
5863 MAIL if the sender used is not empty.
5865 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5868 (c) A new callout option, "use_sender" has been added for people who want
5869 the modified behaviour.
5875 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5878 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5879 indeed breaks things for older releases.
5881 3. Added additional logging to the case where there is a problem reading data
5882 from a filter that is running in a subprocess using a pipe, in order to
5883 try to track down a specific problem.
5885 4. Testing facility fudge: when running in the test harness and attempting
5886 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5887 getting "No route to host". Convert this to a timeout.
5889 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5892 6. Some OS don't have socklen_t but use size_t instead. This affects the
5893 fifth argument of getsockopt() amongst other things. This is now
5894 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5895 can be set for individual OS. I have set it for SunOS5, OSF1, and
5896 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5897 some earlier ones do not.
5899 7. Change 4.30/15 was not doing the test caselessly.
5901 8. The standard form for an IPv6 address literal was being rejected by address
5902 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5903 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5904 this, as well as the form without the "IPv6" on the front (but only when
5905 address literals are enabled, of course).
5907 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5909 10. Exim crashed if a message with an empty sender address specified by -f
5910 encountered a router with an errors_to setting. This could be provoked only
5911 by a command such as
5915 where an empty string was supplied; "<>" did not hit this bug.
5917 11. Installed PCRE release 4.5.
5919 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5920 remained set. It is now erased.
5922 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5923 times from message ids (which are base 36 rather than the normal 62).
5925 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5926 were being counted as actual protocol errors, and logged if the log
5927 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5928 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5929 and DATA following a set of rejected RCPTs do not count as protocol errors.
5930 In other words, Exim assumes they were pipelined, though this may not
5931 actually be the case. Of course, in all cases the client gets an
5932 appropriate error code.
5934 15. If a lookup fails in an ACL condition, a message about the failure may
5935 be available; it is used if testing the ACL cannot continue, because most
5936 such messages specify what the cause of the deferral is. However, some
5937 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5938 that caused an old message to be retained and used if a later statement
5939 caused a defer, replacing the real cause of the deferral.
5941 16. If an IP address had so many PTR records that the DNS lookup buffer
5942 was not large enough to hold them, Exim could crash while trying to process
5943 the truncated data. It now detects and logs this case.
5945 17. Further to 4.21/58, another change has been made: if (and only if) the
5946 first line of a message (the first header line) ends with CRLF, a bare LF
5947 in a subsequent header line has a space inserted after it, so as not to
5948 terminate the header.
5950 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5951 unnecessarily, used atoi() instead of strtol(), and didn't check the
5952 termination when getting file sizes from file names by regex.
5954 19. Completely re-implemented the support for maildirsize files, in the light
5955 of a number of problems with the previous contributed implementation
5956 (4.30/29). In particular:
5958 . If the quota is zero, the maildirsize file is maintained, but no quota is
5961 . If the maildir directory does not exist, it is created before any attempt
5962 to write a maildirsize file.
5964 . The quota value in the file is just a cache; if the quota is changed in
5965 the transport, the new value overrides.
5967 . A regular expression is available for excluding directories from the
5970 20. The autoreply transport checks the characters in options that define the
5971 message's headers; it allows continued headers, but it was checking with
5972 isspace() after an embedded newline instead of explicitly looking for a
5975 21. If all the "regular" hosts to which an address was routed had passed their
5976 expiry times, and had not reached their retry times, the address was
5977 bounced, even if fallback hosts were defined. Now Exim should go on to try
5980 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5981 equivalent code in the SMTP transport. Some hosts send humungous responses
5982 to HELO/EHLO, more than 1024 it seems.
5984 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5985 gives compiler warnings in some environments. I've now done it "properly",
5988 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5989 (because of problems with the built-in one) was declared to return uschar *
5990 instead of char *, causing compiler failure.
5992 25. Fixed a file descriptor leak when processing alias/forward files.
5994 26. Fixed a minor format string issue in dbfn.c.
5996 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5998 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5999 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6002 29. When a sender address is verified, it is cached, to save repeating the test
6003 when there is more than one recipient in a message. However, when the
6004 verification involves a callout, it is possible for different callout
6005 options to be set for different recipients. It is too complicated to keep
6006 track of this in the cache, so now Exim always runs a verification when a
6007 callout is required, relying on the callout cache for the optimization.
6008 The overhead is duplication of the address routing, but this should not be
6011 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6012 to be verified with callout=postmaster, and the main callout worked but the
6013 postmaster check failed, the verification correctly failed. However, if a
6014 subsequent RCPT command asked for sender verification *without* the
6015 postmaster check, incorrect caching caused this verification also to fail,
6018 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6019 it was not caching the DNS options (qualify_single, search_parents) that
6020 were used when the lookup failed. A subsequent lookup with different
6021 options therefore always gave the same answer, though there were cases
6022 where it should not have. (Example: a "domains = !$mx_any" option on a
6023 dnslookup router: the "domains" option is always processed without any
6024 widening, but the router might have qualify_single set.) Now Exim uses the
6025 cached value only when the same options are set.
6027 32. Added John Jetmore's "exipick" utility to the distribution.
6029 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6030 than a timeout (e.g. a certificate is required, and is not provided), an
6031 Exim server now closes the connection immediately. Previously it waited for
6032 the client to close - but if the client is SSL, it seems that they each
6033 wait for each other, leading to a delay before one of them times out.
6035 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6036 maintained 0.8.x compatibility because I don't think many are using it, and
6037 it is clearly obsolete.
6039 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6042 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6043 client certificate was expired. A simple patch fixes this, though I don't
6044 understand the full logic of why the verify callback is called multiple
6047 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6048 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6049 which causes problems with some clients (such as the Certicom SSL Plus
6050 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6051 disables the coutermeasure allowing Eudora to connect."
6053 38. Exim was not checking that a write() to a log file succeeded. This could
6054 lead to Bad Things if a log got too big, in particular if it hit a file
6055 size limit. Exim now panics and dies if it cannot write to a log file, just
6056 as it does if it cannot open a log file.
6058 39. Modified OS/Makefile-Linux so that it now contains
6060 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6062 The two -D definitions ensure that Exim is compiled with large file
6063 support, which makes it possible to handle log files that are bigger than
6066 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6067 instance) a domain was checked against a named list that involved a lookup,
6068 causing $domain_data to be set, then another domain was checked against the
6069 same list, then the first domain was re-checked, the value of $domain_data
6070 after the final check could be wrong. In particular, if the second check
6071 failed, it could be set empty. This bug probably also applied to
6074 41. The strip_trailing_dot option was not being applied to the address given
6075 with the -f command-line option.
6077 42. The code for reading a message's header from the spool was incrementing
6078 $received_count, but never initializing it. This meant that the value was
6079 incorrect (doubled) while delivering a message in the same process in which
6080 it was received. In the most common configuration of Exim, this never
6081 happens - a fresh exec is done - but it can happen when
6082 deliver_drop_privilege is set.
6084 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6085 it now includes up to 150 characters of the unexpected data in the log
6088 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6089 and building data strings. The size of both of these buffers was 10 000
6090 bytes - far larger than anybody would *ever* want, thought I. Needless to
6091 say, somebody hit the limit. I have increased the maximum line length to
6092 20 000 and the maximum data length of concatenated lines to 100 000. I have
6093 also fixed two bugs, because there was no checking on these buffers. Tsk,
6094 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6095 buffer is too small.
6097 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6098 lsearch lookups. Now it does.
6100 46. When parsing a route_list item in a manualroute router, a fixed-length
6101 buffer was used for the list of hosts. I made this 1024 bytes long,
6102 thinking that nobody would ever have a list of hosts that long. Wrong.
6103 Somebody had a whole pile of complicated expansion conditions, and the
6104 string was silently truncated, leading to an expansion error. It turns out
6105 that it is easier to change to an unlimited length (owing to other changes
6106 that have happened since this code was originally written) than to build
6107 structure for giving a limitation error. The length of the item that
6108 expands into the list of hosts is now unlimited.
6110 47. The lsearch lookup could not handle data where the length of text line was
6111 more than 4095 characters. Such lines were truncated, leading to shortened
6112 data being returned. It should now handle lines of any length.
6114 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6115 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6118 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6119 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6120 no longer happens in comment lines. A list of replacements is now placed
6121 at the head of all of the source files, except those whose only change is
6122 to replace PERL_COMMAND in the very first #! line.
6124 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6125 messages on the queue, with a bottom-up merge sort, using code contributed
6126 by Michael Haardt. This should make operations like -bp somewhat faster on
6127 large queues. It won't affect queue runners, except when queue_run_in_order
6130 51. Installed eximstats 1.31 in the distribution.
6132 52. Added support for SRV lookups to the dnslookup router.
6134 53. If an ACL referred to $message_body or $message_body_end, the value was not
6135 reset for any messages that followed in the same SMTP session.
6137 54. The store-handling optimization for building very long strings was not
6138 differentiating between the different store pools. I don't think this
6139 actually made any difference in practice, but I've tidied it.
6141 55. While running the routers to verify a sender address, $sender_address
6142 was still set to the sender address. This is wrong, because when routing to
6143 send a bounce to the sender, it would be empty. Therefore, I have changed
6144 it so that, while verifying a sender address, $sender_address is set to <>.
6145 (There is no change to what happens when verifying a recipient address.)
6147 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6148 target A or AAAA records (if not already returned) without resetting the
6149 qualify_single or search_parents options of the DNS resolver. These are
6150 inappropriate in this case because the targets of MX and SRV records must
6151 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6152 target that, when qualified, matched something in the local domain. These
6153 two options are now turned off when doing these lookups.
6155 57. It seems that at least some releases of Reiserfs (which does not have the
6156 concept of a fixed number of inodes) returns zero and not -1 for the
6157 number of available inodes. This interacted badly with check_spool_inodes,
6158 which assumed that -1 was the "no such thing" setting. What I have done is
6159 to check that the total number of inodes is greater than zero before doing
6160 the test of how many are available.
6162 58. When a "warn" ACL statement has a log_message modifier, the message is
6163 remembered, and not repeated. This is to avoid a lot of repetition when a
6164 message has many recipients that cause the same warning to be written.
6165 However, Exim was preserving the list of already written lines for an
6166 entire SMTP session, which doesn't seem right. The memory is now reset if a
6167 new message is started.
6169 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6170 debugging output unless log_rewrite was also set.
6172 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6173 of (int)(handle) when we know that handle contains (void *)(-1).
6175 61. The Exim daemon panic-logs an error return when it closes the incoming
6176 connection. However "connection reset by peer" seems to be common, and
6177 isn't really an error worthy of noting specially, so that particular error
6180 62. When Exim is trying to find all the local interfaces, it used to panic and
6181 die if the ioctl to get the interface flags failed. However, it seems that
6182 on at least one OS (Solaris 9) it is possible to have an interface that is
6183 included in the list of interfaces, but for which you get a failure error
6184 for this call. This happens when the interface is not "plumbed" into a
6185 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6186 failure of the "get flags" call assumes that the interface is down.
6188 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6189 makes life easier for people who are doing arithmetic on fields extracted
6190 from dates, where you often get leading zeros that should not be
6191 interpreted as octal.
6193 64. Added qualify_domain to the redirect router, to override the global
6196 65. If a pathologically long header line contained very many addresses (the
6197 report of this problem mentioned 10 000) and each of them was rewritten,
6198 Exim could use up a very large amount of memory. (It kept on making new
6199 copies of the header line as it rewrote, and never released the old ones.)
6200 At the expense of a bit more processing, the header rewriting function has
6201 been changed so that it no longer eats memory in this way.
6203 66. The generation of the Received: header has been moved from the time that a
6204 message starts to be received, to the time that it finishes. The timestamp
6205 in the Received: header should now be very close to that of the <= log
6206 line. There are two side-effects of this change:
6208 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6209 logged header lines no longer include the local Received: line, because
6210 it has not yet been created. The same applies to a copy of the message
6211 that is returned to a non-SMTP sender when a message is rejected.
6213 (b) When a filter file is tested using -bf, no additional Received: header
6214 is added to the test message. After some thought, I decided that this
6217 This change does not affect the value of $received_for. It is still set
6218 after address rewriting, but before local_scan() is called.
6220 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6222 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6223 gave an unhelpful panic error message, and a defer error. I have managed to
6224 change this behaviour so that it now rejects any supplied certificate,
6225 which seems right, as the list of acceptable certificates is empty.
6227 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6228 gave an unhelpful defer error. I have not managed to make this reject any
6229 supplied certificates, but the error message it gives is "no certificate
6230 supplied", which is not helpful.
6232 70. exigrep's output now also includes lines that are not associated with any
6233 message, but which match the given pattern. Implemented by a patch from
6234 Martin Sluka, which also tidied up the Perl a bit.
6236 71. Recipient callout verification, like sender verification, was using <> in
6237 the MAIL FROM command. This isn't really the right thing, since the actual
6238 sender may affect whether the remote host accepts the recipient or not. I
6239 have changed it to use the actual sender in the callout; this means that
6240 the cache record is now keyed on a recipient/sender pair, not just the
6241 recipient address. There doesn't seem to be a real danger of callout loops,
6242 since a callout by the remote host to check the sender would use <>.
6243 [SEE ABOVE: changed after hitting problems.]
6245 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6246 temporary errors. However, in the case of such a code being given after
6247 the end of a data transmission (i.e. after ".") Exim was failing to write
6248 a retry record for the message. (Yes, there was some broken host that was
6249 actually sending 8xx at this point.)
6251 73. An unknown lookup type in a host list could cause Exim to panic-die when
6252 the list was checked. (An example that provoked this was putting <; in the
6253 middle of a list instead of at the start.) If this happened during a DATA
6254 ACL check, a -D file could be left lying around. This kind of configuration
6255 error no longer causes Exim to die; instead it causes a defer error. The
6256 incident is still logged to the main and panic logs.
6258 74. Buglet left over from Exim 3 conversion. The message "too many messages
6259 in one connection" was written to the rejectlog but not the mainlog, except
6260 when address rewriting (yes!) was being logged.
6262 75. Added write_rejectlog option.
6264 76. When a system filter was run not as root (that is, when system_filter_user
6265 was set), the values of the $n variables were not being returned to the
6266 main process; thus, they were not subsequently available in the $sn
6269 77. Added +return_path_on_delivery log selector.
6271 78. A connection timeout was being treated differently from recipients deferred
6272 when testing hosts_max_try with a message that was older than the host's
6273 retry timeout. (The host should not be counted, thus allowing all hosts to
6274 be tried at least once before bouncing.) This may have been the cause of an
6275 occasionally reported bug whereby a message would remain on the queue
6276 longer than the retry timeout, but would be bounced if a delivery was
6277 forced. I say "may" because I never totally pinned down the problem;
6278 setting up timeout/retry tests is difficult. See also the next item.
6280 79. The ultimate address timeout was not being applied to errors that involved
6281 a combination of host plus message (for example, a timeout on a MAIL
6282 command). When an address resolved to a number of possible hosts, and they
6283 were not all tried for each delivery (e.g. because of hosts_max_try), a
6284 message could remain on the queue longer than the retry timeout.
6286 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6289 81. Fixed an obscure SMTP outgoing bug which required at least the following
6290 conditions: (a) there was another message waiting for the same server;
6291 (b) the server returned 5xx to all RCPT commands in the first message so
6292 that the message was not completed; (c) the server dropped the connection
6293 or gave a negative response to the RSET that Exim sends to abort the
6294 transaction. The observed case was a dropped connection after DATA that had
6295 been sent in pipelining mode. That is, the server had advertised PIPELINING
6296 but was not implementing it correctly. The effect of the bug was incorrect
6297 behaviour, such as trying another host, and this could lead to a crash.
6303 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6304 and daemon.c were passed as pointers to ints; they should have been
6305 pointers to socklen_t variables (which are typically unsigned ints).
6307 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6310 3. Fixed a really odd bug that affected only the testing scheme; patching a
6311 certain fixed string in the binary changed the value of another string that
6312 happened to be identical to the end of the original first string.
6314 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6315 name", it returns that address as the IP address. On some operating
6316 systems (e.g. Solaris), it also passes back the IP address string as the
6317 "host name". However, on others (e.g. Linux), it passes back an empty
6318 string. Exim wasn't checking for this, and was changing the host name to an
6319 empty string, assuming it had been canonicalized.
6321 5. Although rare, it is permitted to have more than one PTR record for a given
6322 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6323 all the names associated with an address, because they do in Solaris.
6324 However, it seems that they do not in Linux for data that comes from the
6325 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6326 I found this out when I moved to a new Linux workstation and tried to run
6327 the Exim test suite.
6329 To get round this problem I have changed the code so that it now does its
6330 own call to the DNS to look up PTR records when searching for a host name.
6331 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6332 addresses that are only in /etc/hosts are still found.
6334 This behaviour is, however, controlled by an option called host_lookup_
6335 order, which defaults to "bydns:byaddr". If people want to use the other
6336 order, or indeed, just use one or the other means of lookup, they can
6337 specify it in this variable.
6339 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6340 some operating systems, this comes back from gethostbyaddr() as an empty
6341 string, and this is what Exim used to test for. However, it seems that in
6342 other systems, "." is yielded. Exim now tests for this case too.
6344 7. The values of check_spool_space and check_log_space are now held internally
6345 as a number of kilobytes instead of an absolute number of bytes. If a
6346 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6347 kilobyte. This means that much larger values can be stored.
6349 8. Exim monitor: an attempt to get the action menu when not actually pointing
6350 at a message produces an empty menu entitled "No message selected". This
6351 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6352 no entries in it ("Shell widget menu has zero width and/or height"). So I
6353 have added a single, blank menu entry in this case.
6355 9. Added ${quote_local_part.
6357 10. MIME decoding is now applied to the contents of Subject: header lines when
6360 11. Now that a reference to $sender_host_address automatically causes a reverse
6361 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6362 host lookup before query-style lookups in lists that might use this
6363 variable. This has therefore been abolished, and the "net-" prefix is no
6364 longer necessary for query-style lookups.
6366 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6367 have been a typo for LFLAGS, so it has been changed.
6369 13. The install script calls Exim with "-C /dev/null" in order to find the
6370 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6371 to be output. However, since Exim outputs its version number before the
6372 error, it didn't break the script. It just looked ugly. I fixed this by
6373 always allowing "-C /dev/null" if the caller is root.
6375 14. Ignore overlarge ACL variable number when reading spool file - insurance
6376 against a later release with more variables having written the file.
6378 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6379 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6380 this, as well as the form without the "IPv6" on the front.
6382 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6383 OS/Makefile-Darwin file.
6385 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6386 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6388 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6389 variable. However, some OS use ints for inodes. I've added cast to long int
6390 to get rid of the compiler warning.
6392 19. I had forgotten to lock out "/../" in configuration file names when
6393 ALT_CONFIG_PREFIX was set.
6395 20. Routers used for verification do not need to specify transports. However,
6396 if such a router generated a host list, and callout was configured, Exim
6397 crashed, because it could not find a port number from the (non-existent)
6398 transport. It now assumes port 25 in this circumstance.
6400 21. Added the -t option to exigrep.
6402 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6403 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6404 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6405 them is set, LOOKUP_LSEARCH is forced.
6407 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6409 24. Added sender and host information to the "rejected by local_scan()" log
6410 line; previously there was no indication of these.
6412 25. Added .include_if_exists.
6414 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6415 Linux. It turns out that not all file systems support this. Apparently some
6416 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6417 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6418 not supported on the file descriptor, is now ignored when Exim is trying to
6419 sync a directory. This applies only to Linux.
6421 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6423 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6424 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6427 29. Added support for maildirsize files from supplied patch (modified a bit).
6429 30. The use of :fail: followed by an empty string could lead Exim to respond to
6430 sender verification failures with (e.g.):
6432 550 Verification failed for <xxx>
6433 550 Sender verify failed
6435 where the first response line was missing the '-' that indicates it is not
6436 the final line of the response.
6438 31. The loop for finding the name of the user that called Exim had a hardwired
6439 limit of 10; it now uses the value of finduser_retries, which is used for
6440 all other user lookups.
6442 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6445 33. Exim was neglecting to zero errno before one call of strtol() when
6446 expanding a string and expecting an integer value. On some systems this
6447 resulted in spurious "integer overflow" errors. Also, it was casting the
6448 result into an int without checking.
6450 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6451 did not work. The code looks as if it has *never* worked, though it appears
6452 to have been documented since at least release 1.62. I have made it work.
6454 35. The "timeout_DNS" error in retry rules, also documented since at least
6455 1.62, also never worked. As it isn't clear exactly what this means, and
6456 clearly it isn't a major issue, I have abolished the feature by treating it
6457 as "timeout", and writing a warning to the main and panic logs.
6459 36. The display of retry rules for -brt wasn't always showing the error code
6462 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6463 timeout_connect_A, timeout_connect_MX.
6465 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6466 to the empty sender.
6468 39. The daemon was not analysing the content of -oX till after it had closed
6469 stderr and disconnected from the controlling terminal. This meant that any
6470 syntax errors were only noted on the panic log, and the return code from
6471 the command was 0. By re-arranging the code a little, I've made the
6472 decoding happen first, so such errors now appear on stderr, and the return
6473 code is 1. However, the actual setting up of the sockets still happens in
6474 the disconnected process, so errors there are still only recorded on the
6477 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6478 connections (as happens on some IP stacks) was logged at start up time as
6479 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6480 it from "IPv6 and IPv4", which means that two separate sockets are being
6483 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6484 says whether AF_INET or AF_INET6 was passed as an argument.
6486 42. Exiwhat output was messed up when time zones were included in log
6489 43. Exiwhat now gives more information about the daemon's listening ports,
6490 and whether -tls-on-connect was used.
6492 44. The "port" option of the smtp transport is now expanded.
6494 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6495 silently ignored. Now an error message is written to the main and panic
6498 46. There's a new ACL modifier called "logwrite" which writes to a log file
6499 as soon as it is encountered.
6501 47. Added $local_user_uid and $local_user_gid at routing time.
6503 48. Exim crashed when trying to verify a sender address that was being
6506 49. Exim was recognizing only a space character after ".include". It now also
6507 recognizes a tab character.
6509 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6510 extracting the relevant information from the specification. The man page no
6511 longer contains scrambled data for the -d option, and I've added a section
6512 at the front about calling Exim under different names.
6514 51. Added "extra_headers" argument to the "mail" command in filter files.
6516 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6519 53. Installed eximstats 1.29.
6521 54. Added transport_filter_timeout as a generic transport option.
6523 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6524 Cc: header lines. This was required by RFC 822, but it not required by RFC
6527 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6528 incoming messages that did not have them. Now it does so only if the
6529 message originates locally, that is, if there is no associated remote host
6530 address. When Resent- header lines are present, this applies to the Resent-
6531 lines rather than the non-Resent- lines.
6533 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6534 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6536 58. Messages for configuration errors now include the name of the main
6537 configuration file - useful now that there may be more than one file in a
6538 list (.included file names were always shown).
6540 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6541 for those rare installations that do not start the daemon as root or run it
6542 setuid root. I've cut out the call to initgroups() if the daemon is not
6545 60. The Exim user and group can now be bound into the binary as text strings
6546 that are looked up at the start of Exim's processing.
6548 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6550 62. Added $mailstore_basename variable.
6552 63. Installed patch to sieve.c from Michael Haardt.
6554 64. When Exim failed to open the panic log after failing to open the main log,
6555 the original message it was trying to log was written to stderr and debug
6556 output, but if they were not available (the usual case in production), it
6557 was lost. Now it is written to syslog before the two lines that record the
6558 failures to open the logs.
6560 65. Users' Exim filters run in subprocesses under the user's uid. It is
6561 possible for a "deliver" command or an alias in a "personal" command to
6562 provoke an address rewrite. If logging of address rewriting is configured,
6563 this fails because the process is not running as root or exim. There may be
6564 a better way of dealing with this, but for the moment (because 4.30 needs
6565 to be released), I have disabled address rewrite logging when running a
6566 filter in a non-root, non-exim process.
6572 1. The buildconfig auxiliary program wasn't quoting the value set for
6573 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6574 not defined. This bug was masked in 4.22 by the effect that was fixed in
6577 2. Some messages that were rejected after a message id was allocated were
6578 shown as "incomplete" by exigrep. It no longer does this for messages that
6579 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6581 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6582 have allow_domain_literals set, the ID did not get logged in the <= line.
6583 Domain literals are now always recognized in Message-ID: header lines.
6585 4. The first argument for a ${extract expansion item is the key name or field
6586 number. Leading and trailing spaces in this item were not being ignored,
6587 causing some misleading effects.
6589 5. When deliver_drop_privilege was set, single queue runner processes started
6590 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6591 same command in the process it spins off) were not dropping privilege.
6593 6. When the daemon running as "exim" started a queue runner, it always
6594 re-executed Exim in the spun-off process. This is a waste of effort when
6595 deliver_drop_privilege is set. The new process now just calls the
6596 queue-runner function directly.
6602 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6605 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6606 ignored. Though the use of -oP was forcing the writing of a pid file, it
6607 was always written to the default place.
6609 3. If the message "no IP address found for host xxxx" is generated during
6610 incoming verification, it is now followed by identification of the incoming
6611 connection (so you can more easily find what provoked it).
6613 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6615 5. Added some features to "harden" Exim a bit more against certain attacks:
6617 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6618 be put in Local/Makefile. This is like the never_users runtime option,
6619 but it cannot be overridden. The default setting is "root".
6621 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6622 prefix string with which any file named in a -C command line option
6625 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6626 is retained for -C and -D only if the caller of Exim is root. Without
6627 it, the exim user may also use -C and -D and retain privilege.
6629 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6630 command line option is disabled.
6632 6. Macro names set by the -D option must start with an upper case letter, just
6633 like macro names defined in the configuration file.
6635 7. Added "dereference=" facility to LDAP.
6637 8. Two instances of the typo "uknown" in the source files are fixed.
6639 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6640 the Configure-Makefile script screwed up while processing it.
6642 10. Incorporated PCRE 4.4.
6644 11. The SMTP synchronization check was not operating right at the start of an
6645 SMTP session. For example, it could not catch a HELO sent before the client
6646 waited for the greeting. There is now a check for outstanding input at the
6647 point when the greeting is written. Because of the duplex, asynchronous
6648 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6649 way, but not yet received, when the check is performed.
6651 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6652 on TCP/IP sockets, because this apparently causes some broken clients to
6655 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6656 unchanged) from the Cygwin maintainer.
6658 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6659 was supported instead of testing for mbx. Effectively a typo.
6661 15. The spa authenticator server code was not checking that the input it
6662 received was valid base64.
6664 16. The debug output line for the "set" modifier in ACLs was not showing the
6665 name of the variable that was being set.
6667 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6669 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6670 lookup on its own. Something else was needed to trigger the lookup. For
6671 example, a match in host_lookup or the need for a host name in a host list.
6672 Now, if $sender_host_name is referenced and the host name has not yet been
6673 looked up, a lookup is performed. If the lookup fails, the variable remains
6674 empty, and $host_lookup_failed is set to "1".
6676 19. Added "eqi" as a case-independent comparison operator.
6678 20. The saslauthd authentication condition could segfault if neither service
6679 nor realm was specified.
6681 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6682 error message that was logged was misleading, and incoming SMTP
6683 connections were dropped. The message is now more accurate, and temporary
6684 errors are given to SMTP connections.
6686 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6687 (except RSET) with a 421 error, until QUIT is received. However, it was
6688 failing to send a response to QUIT.
6690 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6691 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6692 ACL is now run after the helo_try_verify_hosts code.
6694 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6695 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6696 case-independent, so other case variants are also recognized). Apparently
6697 some systems use these upper case variants.
6699 25. If more than two messages were waiting for the same host, and a transport
6700 filter was specified for the transport, Exim sent two messages over the
6701 same TCP/IP connection, and then failed with "socket operation on non-
6702 socket" when it tried to send the third.
6704 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6706 27. The extern definition of crypt16() in expand.c was not being excluded when
6707 the OS had its own crypt16() function.
6709 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6710 as a preferred synonym for return_size_limit, both as an option and as an
6713 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6715 30. Changed the default configuration ACL to relax the local part checking rule
6716 for addresses that are not in any local domains. For these addresses,
6717 slashes and pipe symbols are allowed within local parts, but the sequence
6718 /../ is explicitly forbidden.
6720 31. SPA server authentication was not clearing the challenge buffer before
6723 32. log_message in a "warn" ACL statement was writing to the reject log as
6724 well as to the main log, which contradicts the documentation and doesn't
6725 seem right (because no rejection is happening). So I have stopped it.
6727 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6728 However, I am unable to do any testing of this.
6730 34. Fixed an infelicity in the appendfile transport. When checking directories
6731 for a mailbox, to see if any needed to be created, it was accidentally
6732 using path names with one or more superfluous leading slashes; tracing
6733 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6735 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6736 contents are added to the log line that is written for every discarded
6737 recipient. (Previously a log_message setting was ignored.)
6739 36. The ${quote: operator now quotes the string if it is empty.
6741 37. The install script runs exim in order to find its version number. If for
6742 some reason other than non-existence or emptiness, which it checks, it
6743 could not run './exim', it was installing it with an empty version number,
6744 i.e. as "exim-". This error state is now caught, and the installation is
6747 38. An argument was missing from the function that creates an error message
6748 when Exim fails to connect to the socket for saslauthd authentication.
6749 This could cause Exim to crash, or give a corrupted message.
6751 39. Added isip, isip4, and isip6 to ${if conditions.
6753 40. The ACL variables $acl_xx are now saved with the message, and can be
6754 accessed later in routers, transports, and filters.
6756 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6757 strings in the file are not string-expanded.
6759 42. If a MAIL command specified a SIZE value that was too large to fit into an
6760 int variable, the check against message_size_limit failed. Such values are
6761 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6762 day this will have to be increased, but I don't think I want to be around
6763 when emails are that large.
6770 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6771 iconv() is not standard in FreeBSD.
6773 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6774 IPv6 enabled. The observed symptom was a segmentation fault on return from
6775 the function os_common_find_running_interfaces() in src/os.c.
6777 3. In the check_special_case() function in daemon.c I had used "errno" as an
6778 argument name, which causes warnings on some systems. This was basically a
6779 typo, since it was named "eno" in the comments!
6781 4. The code that waits for the clock to tick (at a resolution of some fraction
6782 of a second) so as to ensure message-id uniqueness was always waiting for
6783 at least one whole tick, when it could have waited for less. [This is
6784 almost certainly not relevant at current processor speeds, where it is
6785 unlikely to ever wait at all. But we try to future-proof.]
6787 5. The function that sleeps for a time interval that includes fractions of a
6788 second contained a race. It did not block SIGALRM between setting the
6789 timer, and suspending (a couple of lines later). If the interval was short
6790 and the sigsuspend() was delayed until after it had expired, the suspension
6791 never ended. On busy systems this could lead to processes getting stuck for
6794 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6795 process, before it forks any delivery processes. The open lookup caching
6796 mechanism meant that the open file or database connection was passed into
6797 the delivery process. The problem was that delivery processes always tidy
6798 up cached lookup data. This could cause a problem for the next delivery
6799 process started by the queue runner, because the external queue runner
6800 process does not know about the closure. So the next delivery process
6801 still has data in the lookup cache. In the case of a file lookup, there was
6802 no problem because closing a file descriptor in a subprocess doesn't affect
6803 the parent. However, if the lookup was caching a connection to a database,
6804 the connection was closed, and the second delivery process was likely to
6805 see errors such as "PGSQL: query failed: server closed the connection
6806 unexpectedly". The problem has been fixed by closing all cached lookups
6807 in a queue runner before running a delivery process.
6809 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6810 seem to have the "const" qualifier which it has on other OS. I've
6813 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6814 *of the same type* (client or server) with the same public name that an
6815 error should be diagnosed.
6817 9. When Exim looked up a host name for an IP address, but failed to find the
6818 original IP address when looking up the host name (a safety check), it
6819 output the message "<ip address> does not match any IP for NULL", which was
6820 confusing, to say the least. The bug was that the host name should have
6821 appeared instead of "NULL".
6823 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6824 user that is built into the binary, and the -C or -D options is used, root
6825 privilege is dropped before the configuration file is read. In addition,
6826 logging is switched to stderr instead of the normal log files. If the
6827 configuration then re-defines the Exim user, the unprivileged environment
6828 is probably not what is expected, so Exim logs a panic warning message (but
6831 However, if deliver_drop_privilege is set, the unprivileged state may well
6832 be exactly what is intended, so the warning has been cut out in that case,
6833 and Exim is allowed to try to write to its normal log files.
6839 1. smtp_return_error_details was not giving details for temporary sender
6840 or receiver verification errors.
6842 2. Diagnose a configuration error if two authenticators have the same public
6845 3. Exim used not to create the message log file for a message until the first
6846 delivery attempt. This could be confusing when incoming messages were held
6847 for policy or load reasons. The message log file is now created at the time
6848 the message is received, and an initial "Received" line is written to it.
6850 4. The automatically generated man page for command line options had a minor
6851 bug that caused no ill effects; however, a more serious problem was that
6852 the procedure for building the man page automatically didn't always
6853 operate. Consequently, release 4.20 contains an out-of-date version. This
6854 shouldn't happen again.
6856 5. When building Exim with embedded Perl support, the script that builds the
6857 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6858 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6860 6. The freeze_tell option was not being used for messages that were frozen on
6861 arrival, either by an ACL or by local_scan().
6863 7. Added the smtp_incomplete_transaction log selector.
6865 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6866 was accepting AUTH without a new EHLO.
6868 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6869 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6872 10. Logging of TCP/IP connections (when configured) now happens in the main
6873 daemon process instead of the child process, so that the TCP/IP connection
6874 count is more accurate (but it can never be perfect).
6876 11. The use of "drop" in a nested ACL was not being handled correctly in the
6877 outer ACL. Now, if condition failure induced by the nested "drop" causes
6878 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6879 or "require"), the connection is dropped.
6881 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6882 that yield "discard" can now be used with an "accept" or a "discard" verb,
6883 but an error is generated for any others (because I can't see a useful way
6884 to define what should happen).
6886 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6887 are now processed in the same way as lines in the Exim configuration file.
6888 In particular, continuation lines are supported.
6890 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6892 15. Added -ti meaning -t -i.
6894 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6895 domains, and warn by logging if others are found.
6897 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6898 returned from GIFCONF when Exim is trying to find the list of interfaces on
6899 a host. The code in os.c has been modified to copy each ifreq to an aligned
6900 structure in all cases.
6902 Also, in some cases, the returned ifreq's were being copied to a 'struct
6903 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6904 means the last couple of bytes of an IPv6 address could be chopped if the
6905 ifreq contained only a normal sockaddr (14 bytes storage).
6907 18. Named domain lists were not supported in the hosts_treat_as_local option.
6908 An entry such as +xxxx was not recognized, and was treated as a literal
6911 19. Ensure that header lines added by a DATA ACL are included in the reject log
6912 if the ACL subsequently rejects the message.
6914 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6915 MD5 (which is deprecated).
6917 21. When testing a filter file using -bf, Exim was writing a message when it
6918 took the sender from a "From " line in the message, but it was not doing so
6919 when it took $return_path from a Return-Path: header line. It now does.
6921 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6922 with a valid header line field name (a series of printing characters
6923 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6925 23. Changed "disc" in the source to "disk" to conform to the documentation and
6926 the book and for uniformity.
6928 24. Ignore Sendmail's -Ooption=value command line item.
6930 25. When execve() failed while trying to run a command in a pipe transport,
6931 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6932 could be confused with a return value of 69 from the command itself. This
6933 has been changed to 127, the value the shell returns if it is asked to run
6934 a non-existent command. The wording for the related log line suggests a
6935 non-existent command as the problem.
6937 26. If received_header_text expands to an empty string, do not add a Received:
6938 header line to the message. (Well, it adds a token one on the spool, but
6939 marks it "old" so that it doesn't get used or transmitted.)
6941 27. Installed eximstats 1.28 (addition of -nt option).
6943 28. There was no check for failure on the call to getsockname() in the daemon
6944 code. This can fail if there is a shortage of resources on the system, with
6945 ENOMEM, for example. A temporary error is now given on failure.
6947 29. Contrary to the C standard, it seems that in some environments, the
6948 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6949 program. Exim now does this explicitly; it affects the formatting of
6950 timestamps using strftime().
6952 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6953 complaints. I've now initialized all the variables, to avoid this.
6955 32. Header lines added by a system filter were not being "seen" during
6956 transport-time rewrites.
6958 33. The info_callback() function passed to OpenSSL is set up with type void
6959 (*)(SSL *, int, int), as described somewhere. However, when calling the
6960 function (actually a macro) that sets it up, the type void(*)() is
6961 expected. I've put in a cast to prevent warnings from picky compilers.
6963 34. If a DNS black list lookup found a CNAME record, but there were no A
6964 records associated with the domain it pointed at, Exim crashed.
6966 35. If a DNS black list lookup returned more than one A record, Exim ignored
6967 all but the first. It now scans all returned addresses if a particular IP
6968 value is being sought. In this situation, the contents of the
6969 $dnslist_value variable are a list of all the addresses, separated by a
6972 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6973 to accept a host name and all its aliases if the forward lookup for any of
6974 them yielded the IP address of the incoming connection. Now it accepts only
6975 those names whose forward lookup yields the correct IP address. Any other
6976 names are discarded. This closes a loophole whereby a rogue DNS
6977 administrator could create reverse DNS records to break through a
6978 wildcarded host restriction in an ACL.
6980 37. If a user filter or a system filter that ran in a subprocess used any of
6981 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6982 the wrong values were passed to the pipe command ($thisaddress had the
6983 value of $0, $0 had the value of $1, etc). This bug was introduced by
6984 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6986 38. Improved the line breaking for long SMTP error messages from ACLs.
6987 Previously, if there was no break point between 40 and 75 characters, Exim
6988 left the rest of the message alone. Two changes have been made: (a) I've
6989 reduced the minimum length to 35 characters; (b) if it can't find a break
6990 point between 35 and 75 characters, it looks ahead and uses the first one
6991 that it finds. This may give the occasional overlong line, but at least the
6992 remaining text gets split now.
6994 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6995 file descriptors might be low, and that setting 1000 would always raise it.
6996 It turns out that in some environments, the limit is already over 1000 and
6997 that lowering it causes trouble. So now Exim takes care not to decrease it.
6999 40. When delivering a message, the value of $return_path is set to $sender_
7000 address at the start of routing (routers may change the value). By an
7001 oversight, this default was not being set up when an address was tested by
7002 -bt or -bv, which affected the outcome if any router or filter referred to
7005 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7006 log only when "message" or "log_message" are set. However, if one of the
7007 conditions was an address verification, or a call to a nested ACL, the
7008 messages generated by the underlying test were being passed through. This
7009 no longer happens. The underlying message is available in $acl_verify_
7010 message for both "message" and "log_message" expansions, so it can be
7011 passed through if needed.
7013 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7014 new expansion $bh_ to give the encoded byte string without charset
7015 translation. Translation happens only if iconv() is available; HAVE_ICONV
7016 indicates this at build time. HEADERS_CHARSET gives the charset to
7017 translate to; headers_charset can change it in the configuration, and
7018 "headers charset" can change it in an individual filter file.
7020 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7021 that creates RFC 2047 encoded "words" labels them as that charset instead
7022 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7023 expansion operator; (ii) when Exim creates a From: line for a local
7024 message; (iii) when a header line is rewritten to include a "phrase" part.
7026 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7027 buggy, causing it to skip the first lines of messages whose message ID
7028 ended in 'D'. This would not have bitten before Exim release 4.14, because
7029 message IDs were unlikely to end in 'D' before then. The effect was to have
7030 incorrect size information for certain domains.
7032 45. #include "config.h" was missing at the start of the crypt16.c module. This
7033 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7036 46. If there was a timeout during a "random" callout check, Exim treated it as
7037 a failure of the random address, and carried on sending RSET and the real
7038 address. If the delay was just some slowness somewhere, the response to the
7039 original RCPT would be taken as a response to RSET and so on, causing
7040 mayhem of various kinds.
7042 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7043 when I implemented it. It didn't allow for the fact that some option values
7044 may legitimately be negative (e.g. size_addition), and it didn't even do
7045 the right test for positive values.
7047 48. Domain names in DNS records are case-independent. Exim always looks them up
7048 in lower case. Some resolvers return domain names in exactly the case they
7049 appear in the zone file, that is, they may contain uppercase letters. Not
7050 all resolvers do this - some return always lower case. Exim was treating a
7051 change of case by a resolver as a change of domain, similar to a widening
7052 of a domain abbreviation. This triggered its re-routing code and so it was
7053 trying to route what was effectively the same domain again. This normally
7054 caused routing to fail (because the router wouldn't handle the domain
7055 twice). Now Exim checks for this case specially, and just changes the
7056 casing of the domain that it ultimately uses when it transmits the message
7059 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7062 50. If a filter generated a file delivery with a non-absolute name (possible if
7063 no home directory exists for the router), the forbid_file option was not
7066 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7067 the existing equality matching.
7069 52. Exim was using ints instead of ino_t variables in some places where it was
7070 dealing with inode numbers.
7072 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7073 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7074 called TMPDIR, and if it finds it is different, it changes its value.
7076 54. The smtp_printf() function is now made available to local_scan() so
7077 additional output lines can be written before returning. There is also an
7078 smtp_fflush() function to enable the detection of a dropped connection.
7079 The variables smtp_input and smtp_batched_input are exported to
7082 55. Changed the default runtime configuration: the message "Unknown user"
7083 has been removed from the ACL, and instead placed on the localuser router,
7084 using the cannot_route_message feature. This means that any verification
7085 failures that generate their own messages won't get overridden. Similarly,
7086 the "Unrouteable address" message that was in the ACL for unverifiable
7087 relay addresses has also been removed.
7089 56. Added hosts_avoid_esmtp to the smtp transport.
7091 57. The exicyclog script was not checking for the esoteric option
7092 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7093 will work only if exicyclog is run under the appropriate euid.
7095 58. Following a discussion on the list, the rules by which Exim recognises line
7096 endings on incoming messages have been changed. The -dropcr and drop_cr
7097 options are now no-ops, retained only for backwards compatibility. The
7098 following line terminators are recognized: LF CRLF CR. However, special
7099 processing applies to CR:
7101 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7102 nor a local message in the state where . is a terminator.
7104 (ii) If a bare CR is encountered in a header line, an extra space is added
7105 after the line terminator so as not to end the header. The reasoning
7106 behind this is that bare CRs in header lines are most likely either
7107 to be mistakes, or people trying to play silly games.
7109 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7110 was being incorrectly given as 18 bytes larger than it should have been.
7111 This is a VOB (very old bug).
7113 60. This may never have affected anything current, but just in case it has:
7114 When the local host is found other than at the start of a list of hosts,
7115 the local host, those with the same MX, and any that follow, are discarded.
7116 When the list in question was part of a longer list of hosts, the following
7117 hosts (not currently being processed) were also being discarded. This no
7118 longer happens. I'm not sure if this situation could ever has previously
7121 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7124 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7125 header. This is something that is recommended in a new Internet Draft, and
7126 is something that is documented as being done by Sendmail. There are two
7127 possible values. For messages generated by the autoreply transport, Exim
7130 Auto-Submitted: auto-replied
7132 whereas for all other generated messages (e.g. bounces) it adds
7134 Auto-Submitted: auto-generated
7136 63. The "personal" condition in filters now includes a test for the
7137 Auto-Submitted: header. If it contains the string "auto-" the message it
7138 not considered personal.
7140 64. Added rcpt_include_affixes as a generic transport option.
7142 65. Added queue_only_override (default true).
7144 66. Added the syslog_duplication option.
7146 67. If what should have been the first header line of a message consisted of
7147 a space followed by a colon, Exim was mis-interpreting it as a header line.
7148 It isn't of course - it is syntactically invalid and should therefore be
7149 treated as the start of the message body. The misbehaviour could have
7150 caused a number of strange effects, including loss of data in subsequent
7151 header lines, and spool format errors.
7153 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7154 client host had authenticated. This control can now be exercised by an ACL
7155 for more flexibility.
7157 69. By default, callouts do not happen when testing with -bh. There is now a
7158 variant, -bhc, which does actually run the callout code, including
7159 consulting and updating the callout cache.
7161 70. Added support for saslauthd authentication, courtesy of Alexander
7164 71. If statvfs() failed on the spool or log directories while checking their
7165 size for availability, Exim confusingly gave the error "space shortage".
7166 Furthermore, in debugging mode it crashed with a floating point exception.
7167 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7168 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7169 serious problem, Exim now writes to the main and panic logs when this
7170 happens, with details of the failure. It then refuses to accept the
7171 incoming message, giving the message "spool directory problem" or "log
7172 directory problem" with a 421 code for SMTP messages.
7174 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7175 0, 1, and 2 exist, because some OS complain for execs without them (see
7176 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7177 descriptors. However, the code omitted to check that the open succeeded,
7178 causing mysterious errors if for some reason the permissions on /dev/null
7179 got screwed. Now Exim writes a message to the main and panic logs, and
7180 bombs out if it can't open /dev/null.
7182 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7183 interact so that it is all more flexible. It is supposed to remain
7184 backwards compatible. Also added extra_local_interfaces.
7186 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7187 to bomb out with an assertion failure - to the client this appears as a
7188 connection drop. This problem occurs in the part of the code that was taken
7189 from the Samba project. Fortunately, the assertion is in a very simple
7190 function, so I have fixed this by reproducing the function inline in the
7191 one place where it is called, and arranging for authentication to fail
7192 instead of killing the process with assert().
7194 75. The SPA client code was not working when the server requested OEM rather
7195 than Unicode encoding.
7197 76. Added code to make require_files with a specific uid setting more usable in
7198 the case where statting the file as root fails - usually a non-root-mounted
7199 NFS file system. When this happens and the failure is EACCES, Exim now
7200 forks a subprocess and does the per-uid checking as the relevant uid.
7202 77. Added process_log_path.
7204 78. If log_file_path was not explicitly set, a setting of check_log_space or
7205 check_log_inodes was ignored.
7207 79. If a space check for the spool or log partitions fails, the incident is now
7208 logged. Of course, in the latter case the data may get lost...
7210 80. Added the %p formatting code to string_format() so that it can be used to
7211 print addresses in debug_print(). Adjusted all the address printing in the
7212 debugging in store.c to use %p rather than %d.
7214 81. There was a concern that a line of code in smtp_in.c could overflow a
7215 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7216 initially expressed, the concern was not well-founded, because trailing
7217 spaces are removed early. However, if the trailing spaces were followed by
7218 a NULL, they did not get removed, so the overflow was possible. Two fixes
7221 (a) I re-wrote the offending code in a cleaner fashion.
7222 (b) If an incoming SMTP command contains a NULL character, it is rejected
7225 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7226 runs initgroups(), so that if the Exim user is in any additional groups,
7227 they will be used during message reception.
7233 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob