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
53 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
54 When redis_servers is set to a list of > 1 element, and the Redis servers
55 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
56 case of MOVED into a DEFER case instead, thus moving the query onto the
57 next server in the list. For a cluster of N elements, all N servers must
58 be defined in redis_servers.
60 GF/02 Catch and remove uninitialized value warning in exiqsumm
61 Check for existence of @ARGV before looking at $ARGV[0]
63 JH/01 Replace the store_release() internal interface with store_newblock(),
64 which internalises the check required to safely use the old one, plus
65 the allocate and data copy operations duplicated in both (!) of the
68 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
69 modifier. This matches the restriction on the commandline.
71 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
72 Previously only the last row was returned.
74 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
75 we assumed that tags in the header were well-formed, and parsed the
76 element content after inspecting only the first char of the tag.
77 Assumptions at that stage could crash the receive process on malformed
80 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
81 While running the DKIM ACL we operate on the Permanent memory pool so that
82 variables created with "set" persist to the DATA ACL. Also (at any time)
83 DNS lookups that fail create cache records using the Permanent pool. But
84 expansions release any allocations made on the current pool - so a dnsdb
85 lookup expansion done in the DKIM ACL releases the memory used for the
86 DNS negative-cache, and bad things result. Solution is to switch to the
87 Main pool for expansions.
88 While we're in that code, add checks on the DNS cache during store_reset,
89 active in the testsuite.
90 Problem spotted, and debugging aided, by Wolfgang Breyha.
92 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
93 When none of the hosts presented to a transport match an already-open
94 connection, close it and proceed with the list. Previously we would
95 queue the message. Spotted by Lena with Yahoo, probably involving
98 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
99 Previously a spurious "250 OK id=" response was appended to the proper
102 JH/08 The "support for" informational output now, which built with Content
103 Scanning support, has a line for the malware scanner interfaces compiled
104 in. Interface can be individually included or not at build time.
106 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
107 by the template makefile "src/EDITME". The "STREAM" support for an older
108 ClamAV interface method is removed.
110 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
111 rows affected is given instead).
113 JH/11 The runtime Berkeley DB library version is now additionally output by
114 "exim -d -bV". Previously only the compile-time version was shown.
116 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
117 SMTP connection. Previously, when one had more receipients than the
118 first, an abortive onward connection was made. Move to full support for
119 multiple onward connections in sequence, handling cutthrough connection
120 for all multi-message initiating connections.
122 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
123 routers. Previously, a multi-recipient message would fail to match the
124 onward-connection opened for the first recipient, and cause its closure.
126 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
127 a timeout on read on a GnuTLS initiating connection, resulting in the
128 initiating connection being dropped. This mattered most when the callout
129 was marked defer_ok. Fix to keep the two timeout-detection methods
132 JH/15 Relax results from ACL control request to enable cutthrough, in
133 unsupported situations, from error to silently (except under debug)
134 ignoring. This covers use with PRDR, frozen messages, queue-only and
137 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
139 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
140 metadata, resulting in a crash in free().
142 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
143 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
144 Broken also in d185889f4, with init system revamp.
146 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
147 Previously we abruptly closed the connection after reading a malware-
148 found indication; now we go on to read the "scan ok" response line,
151 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
152 ACL. Previously, a crash would result.
154 JH/19 Speed up macro lookups during configuration file read, by skipping non-
155 macro text after a replacement (previously it was only once per line) and
156 by skipping builtin macros when searching for an uppercase lead character.
158 JH/20 DANE support moved from Experimental to mainline. The Makefile control
159 for the build is renamed.
161 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
162 was allocated for every new TLS startup, meaning one per message. Fix
163 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
165 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
166 reported the original. Fix to report (as far as possible) the ACL
167 result replacing the original.
169 JH/23 Fix memory leak during multi-message connections using STARTTLS under
170 OpenSSL. Certificate information is loaded for every new TLS startup,
171 and the resources needed to be freed.
173 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
175 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
178 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
179 DATA response info to the (existing) per-recipient response info for
180 the "C=" log element. It can have useful tracking info from the
181 destination system. Patch from Simon Arlott.
183 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
184 length value. Previously this would segfault.
186 HS/02 Support Avast multiline protoocol, this allows passing flags to
187 newer versions of the scanner.
189 JH/28 Ensure that variables possibly set during message acceptance are marked
190 dead before release of memory in the daemon loop. This stops complaints
191 about them when the debug_store option is enabled. Discovered specifically
192 for sender_rate_period, but applies to a whole set of variables.
193 Do the same for the queue-runner and queue-list loops, for variables set
194 from spool message files. Do the same for the SMTP per-message loop, for
195 certain variables indirectly set in ACL operations.
197 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
198 as a multi-recipient message from a mailinglist manager). The coding had
199 an arbitrary cutoff number of characters while checking for more input;
200 enforced by writing a NUL into the buffer. This corrupted long / fast
201 input. The problem was exposed more widely when more pipelineing of SMTP
202 responses was introduced, and one Exim system was feeding another.
203 The symptom is log complaints of SMTP syntax error (NUL chars) on the
204 receiving system, and refused recipients seen by the sending system
205 (propating to people being dropped from mailing lists).
206 Discovered and pinpointed by David Carter.
208 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
209 replaced by the ${authresults } expansion.
211 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
213 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
214 allows proper process termination in container environments.
216 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
217 Previously the "final dot" had a newline after it; ensure it is CR,LF.
219 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
220 and "err_perm", deprecated since 4.83 when the RFC-defined words
221 "temperror" and "permerror" were introduced.
223 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
224 transport-filters or DKIM-signing. The restriction was lost in the
225 consolidation of verify-callout and delivery SMTP handling.
226 Extend the restriction to also cover ARC-signing.
228 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
229 in defer=pass mode supply a 450 to the initiator. Previously the message
232 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
233 tls_require_ciphers is used as before.
235 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
236 "pass_unscanned". Only tmpfails from the scanner are written to
237 the paniclog, as they may require admin intervention (permission
238 denied, license issues). Other scanner errors (like decompression
239 bombs) do not cause a paniclog entry.
241 JH/36 Fix reinitialisation of DKIM logging variable between messages.
242 Previously it was possible to log spurious information in receive log
245 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
246 triggered odd behaviour from Outlook Express clients.
248 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
251 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
252 since the IETF WG has not yet settled on that versus the original
253 "bare" representation.
255 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
256 Previously the millisecond value corrupted the output.
257 Fix also for syslog_pid=no and log_selector +pid, for which the pid
258 corrupted the output.
264 JH/01 Rework error string handling in TLS interface so that the caller in
265 more cases is responsible for logging. This permits library-sourced
266 string to be attached to addresses during delivery, and collapses
267 pairs of long lines into single ones.
269 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
270 during configuration. Wildcards are allowed and expanded.
272 JH/02 Rework error string handling in DKIM to pass more info back to callers.
273 This permits better logging.
275 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
276 do not close it down and have the child transport start it up again on
277 the passed-on TCP connection. Instead, proxy the child (and any
278 subsequent ones) for TLS via a unix-domain socket channel. Logging is
279 affected: the continued delivery log lines do not have any DNSSEC, TLS
280 Certificate or OCSP information. TLS cipher information is still logged.
282 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
283 identical IP addresses on different listening ports. Will also affect
286 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
287 add noisy ifdef guards to special-case this sillyness.
288 Patch from Bernd Kuhls.
290 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
291 than 255 are no longer allowed.
293 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
294 Disable the session-cache too, which might reduce our load. Since we
295 currrectly use a new context for every connection, both as server and
296 client, there is no benefit for these.
297 GnuTLS appears to not support tickets server-side by default (we don't
298 call gnutls_session_ticket_enable_server()) but client side is enabled
299 by default on recent versions (3.1.3 +) unless the PFS priority string
302 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
303 <https://reproducible-builds.org/specs/source-date-epoch/>.
305 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
306 the check for any unsuccessful recipients did not notice the limit, and
307 erroneously found still-pending ones.
309 JH/08 Pipeline CHUNKING command and data together, on kernels that support
310 MSG_MORE. Only in-clear (not on TLS connections).
312 JH/09 Avoid using a temporary file during transport using dkim. Unless a
313 transport-filter is involved we can buffer the headers in memory for
314 creating the signature, and read the spool data file once for the
315 signature and again for transmission.
317 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
318 4.77 as the kernel support then wasn't solid, having issues in 64bit
319 mode. Now, it's been long enough. Add support for FreeBSD also.
321 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
322 case where the routing stage had gathered several addresses to send to
323 a host before calling the transport for the first, we previously failed
324 to close down TLS in the old transport process before passing the TCP
325 connection to the new process. The new one sent a STARTTLS command
326 which naturally failed, giving a failed delivery and bloating the retry
327 database. Investigation and fix prototype from Wolfgang Breyha.
329 JH/12 Fix check on SMTP command input synchronisation. Previously there were
330 false-negatives in the check that the sender had not preempted a response
331 or prompt from Exim (running as a server), due to that code's lack of
332 awareness of the SMTP input buffering.
334 PP/04 Add commandline_checks_require_admin option.
335 Exim drops privileges sanely, various checks such as -be aren't a
336 security problem, as long as you trust local users with access to their
337 own account. When invoked by services which pass untrusted data to
338 Exim, this might be an issue. Set this option in main configuration
339 AND make fixes to the calling application, such as using `--` to stop
342 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
343 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
344 responses to those, into a single TLS record each way (this usually means
345 a single packet). As a side issue, smtp_enforce_sync now works on TLS
348 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
349 affects you only if you're dancing at the edge of the param size limits.
350 If you are, and this message makes sense to you, then: raise the
351 configured limit or use OpenSSL 1.1. Nothing we can do for older
354 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
355 cmdline element to get the documented default one. Previously it was
358 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
359 are unsuable for non-message-specific reasons - eg. connection timeout,
360 banner-time rejection.
362 JH/16 Fix logging of delivery remote port, when specified by router, under
365 PP/06 Repair manualroute's ability to take options in any order, even if one
366 is the name of a transport.
369 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
371 JH/17 Change the list-building routines interface to use the expanding-string
372 triplet model, for better allocation and copying behaviour.
374 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
375 Previously it was constructed the first time a possibly-matching string
376 was met in the configuration file input during startup; now it is done
379 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
380 compatible one, to avoid the (poorly documented) possibility of a config
381 file in the working directory redirecting the DB files, possibly correpting
382 some existing file. CVE-2017-10140 assigned for BDB.
384 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
385 cache-hot. Previously, although the result was properly cached, the
386 initial verify call returned a defer.
388 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
389 the main verify for receipient in uncached-mode.
391 JH/22 Retire historical build files to an "unsupported" subdir. These are
392 defined as "ones for which we have no current evidence of testing".
394 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
395 if present. Previously it was ignored.
397 JH/24 Start using specified-initialisers in C structure init coding. This is
398 a C99 feature (it's 2017, so now considered safe).
400 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
401 if was a fixed-sized field and bitmask ops via macros; it is now more
404 PP/07 GitHub PR 56: Apply MariaDB build fix.
405 Patch provided by Jaroslav Škarvada.
407 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
408 during Coverity cleanups [4.87 JH/47]
409 Diagnosis and fix provided by Michael Fischer v. Mollard.
411 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
412 the right size to place the terminating semicolon on its own folded
413 line, the header hash was calculated to an incorrect value thanks to
414 the (relaxed) space the fold became.
416 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
417 and confused the parent.
419 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
420 which could crash as a result. This could lead to undeliverable messages.
422 JH/28 Logging: "next input sent too soon" now shows where input was truncated
425 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
426 matters on fast-turnover and PID-randomising systems, which were getting
427 out-of-order delivery.
429 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
430 a possibly-overlapping copy. The symptom was that "Remote host closed
431 connection in response to HELO" was logged instead of the actual 4xx
434 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
435 Previously only that bufferd was discarded, resulting in SYMTP command
438 JH/32 DKIM: when a message has multiple signatures matching an identity given
439 in dkim_verify_signers, run the dkim acl once for each. Previously only
440 one run was done. Bug 2189.
442 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
443 "panic the current process" to "deliberately defer". The panic log is
444 still written with the problem list name; the mail and reject logs now
445 get a temp-reject line for the message that was being handled, saying
446 something like "domains check lookup or other defer". The SMTP 451
447 message is still "Temporary local problem".
449 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
450 A crafted sequence of BDAT commands could result in in-use memory beeing
451 freed. CVE-2017-16943.
453 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
454 from SMTP input. Previously it was always done; now only done for DATA
455 and not BDAT commands. CVE-2017-16944.
457 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
458 to the message (such as an overlong header line). Previously this was
459 not done and we did not exit BDAT mode. Followon from the previous item
460 though a different problem.
466 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
467 than -2003 did; needs libidn2 in addition to libidn.
469 JH/02 The path option on a pipe transport is now expanded before use.
471 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
472 Patch provided by "Björn", documentation fix added too.
474 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
475 missing a wire-to-host endian conversion.
477 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
478 close after a BDAT command line could be taken as a following command,
479 giving a synch failure. Fix by only checking for synch immediately
480 before acknowledging the chunk.
482 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
483 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
484 macro. Patches provided by Josh Soref.
486 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
487 Previously we did not; the RFC seems ambiguous and VRFY is not listed
488 by IANA as a service extension. However, John Klensin suggests that we
491 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
492 the dkim code may be unix-mode line endings rather than smtp wire-format
493 CRLF, so prepend a CR to any bare LF.
495 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
496 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
498 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
499 the first were themselves being wrongly included in the feed into dkim
500 processing; with most chunk sizes in use this resulted in an incorrect
501 body hash calculated value.
503 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
504 DKIM signature block, for verification. Although advised against by
505 standards it is specifically not ruled illegal.
507 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
509 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
510 missing a body hash (the bh= tag).
512 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
513 It seems that HAProxy sends the Proxy Protocol information in clear and
514 only then does a TLS startup, so do the same.
516 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
517 TCP connections (such as for Spamd) unless the daemon successfully set
518 Fast Open mode on its listening sockets. This fixes breakage seen on
519 too-old kernels or those not configured for Fast Open, at the cost of
520 requiring both directions being enabled for TFO, and TFO never being used
521 by non-daemon-related Exim processes.
523 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
524 endings, at least on the first header line. Try to canonify any that get
525 past that check, despite the cost.
527 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
528 now limited to an arbitrary five deep, while parsing addresses with the
529 strip_excess_angle_brackets option enabled.
531 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
532 instead leave the unprompted TLS handshake in socket buffer for the
533 TLS library to consume.
535 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
537 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
539 JH/16 Drop variables when they go out of scope. Memory management drops a whole
540 region in one operation, for speed, and this leaves assigned pointers
541 dangling. Add checks run only under the testsuite which checks all
542 variables at a store-reset and panics on a dangling pointer; add code
543 explicitly nulling out all the variables discovered. Fixes one known
544 bug: a transport crash, where a dangling pointer for $sending_ip_address
545 originally assigned in a verify callout, is re-used.
547 PP/06 Drop '.' from @INC in various Perl scripts.
549 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
551 PP/08 Reduce a number of compilation warnings under clang; building with
552 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
553 should be warning-free.
555 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
557 HS/01 Fix portability problems introduced by PP/08 for platforms where
558 realloc(NULL) is not equivalent to malloc() [SunOS et al].
560 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
561 chunk. This allows us to accept broken chunked messages. We need a more
562 general solution here.
564 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
565 already-broken messages in the queue.
567 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
569 JH/19 Fix reference counting bug in routing-generated-address tracking.
575 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
576 supports it and a size is available (ie. the sending peer gave us one).
578 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
579 years of being deprecated). The replacements are the ACLs
580 acl_smtp_mime and acl_not_smtp_mime.
582 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
583 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
584 or even an in-clear connection were permitted. Now, if the host lookup
585 was dnssec and dane was requested then the host is only used if the
586 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
587 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
588 if one fails this test.
589 This means that a poorly-configured remote DNS will make it incommunicado;
590 but it protects against a DNS-interception attack on it.
592 JH/04 Bug 1810: make continued-use of an open smtp transport connection
593 non-noisy when a race steals the message being considered.
595 JH/05 If main configuration option tls_certificate is unset, generate a
596 self-signed certificate for inbound TLS connections.
598 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
599 in rewrites and routers.
601 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
602 and logged a warning sing 4.83; now they are a configuration file error.
604 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
605 (lacking @domain). Apply the same qualification processing as RCPT.
607 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
609 JH/10 Support ${sha256:} applied to a string (as well as the previous
612 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
613 a cutthrough deliver is pending, as we always want to make a connection.
614 This also avoids re-routing the message when later placing the cutthrough
615 connection after a verify cache hit.
616 Do not update it with the verify result either.
618 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
619 when routing results in more than one destination address.
621 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
622 signing (which inhibits the cutthrough capability). Previously only
623 the presence of an option was tested; now an expansion evaluating as
624 empty is permissible (obviously it should depend only on data available
625 when the cutthrough connection is made).
627 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
628 the relevant preceding SMTP command did not note the pipelining mode.
630 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
631 Previously they were not counted.
633 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
634 as one having no matching records. Previously we deferred the message
635 that needed the lookup.
637 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
638 distinguished as "(=".
640 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
641 for missing MX records. Previously it only worked for missing A records.
643 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
645 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
646 after the data-go-ahead and data-ack. Patch from Jason Betts.
648 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
649 even for a "none" policy. Patch from Tony Meyer.
651 JH/22 Fix continued use of a connection for further deliveries. If a port was
652 specified by a router, it must also match for the delivery to be
655 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
656 When one of the recipients of a message was unsuitable for the connection
657 (has no matching addresses), we lost track of needing to mark it
658 deferred. As a result mail would be lost.
660 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
662 JH/25 Decoding ACL controls is now done using a binary search; the source code
663 takes up less space and should be simpler to maintain. Merge the ACL
664 condition decode tables also, with similar effect.
666 JH/26 Fix problem with one_time used on a redirect router which returned the
667 parent address unchanged. A retry would see the parent address marked as
668 delivered, so not attempt the (identical) child. As a result mail would
671 JH/27 Fix a possible security hole, wherein a process operating with the Exim
672 UID can gain a root shell. Credit to http://www.halfdog.net/ for
673 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
676 JH/28 Enable {spool,log} filesystem space and inode checks as default.
677 Main config options check_{log,spool}_{inodes,space} are now
678 100 inodes, 10MB unless set otherwise in the configuration.
680 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
681 Previously it only applied to the main-section connection policy
684 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
686 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
687 by me. Added RFC7919 DH primes as an alternative.
689 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
690 are not in the system include path.
692 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
693 GnuTLS, when a session startup failed (eg because the client disconnected)
694 Exim did stdio operations after fclose. This was exposed by a recent
695 change which nulled out the file handle after the fclose.
697 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
698 signed directly by the cert-signing cert, rather than an intermediate
699 OCSP-signing cert. This is the model used by LetsEncrypt.
701 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
703 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
704 an incoming connection.
706 HS/02 Bug 1802: Do not half-close the connection after sending a request
709 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
710 fallback to "prime256v1".
712 JH/34 SECURITY: Use proper copy of DATA command in error message.
713 Could leak key material. Remotely exploitable. CVE-2016-9963.
719 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
720 and 3.4.4 - once the server is enabled to respond to an OCSP request
721 it does even when not requested, resulting in a stapling non-aware
722 client dropping the TLS connection.
724 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
725 support variable-length bit vectors. No functional change.
727 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
728 The I= interface field on outgoing lines is now after the H= remote
729 host field, same as incoming lines. There is a separate
730 outgoing_interface log selector which allows you to disable the
733 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
734 If not running log_selector +smtp_connection the mainlog would be held
735 open indefinitely after a "too many connections" event, including to a
736 deleted file after a log rotate. Leave the per net connection logging
737 leaving it open for efficiency as that will be quickly detected by the
738 check on the next write.
740 HS/01 Bug 1671: Fix post transport crash.
741 Processing the wait-<transport> messages could crash the delivery
742 process if the message IDs didn't exist for some reason. When
743 using 'split_spool_directory=yes' the construction of the spool
744 file name failed already, exposing the same netto behaviour.
746 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
747 mime_regex ACL conditions.
749 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
750 to DSN fail messages (bounces): remote IP, remote greeting, remote response
751 to HELO, local diagnostic string.
753 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
754 log line to debug. Even when configured with a tls authenticator many
755 client connections are expected to not authenticate in this way, so
756 an authenticate fail is not an error.
758 HS/02 Add the Exim version string to the process info. This way exiwhat
759 gives some more detail about the running daemon.
761 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
762 matter for fast-change records such as DNSBLs.
764 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
765 retry record, even if constant. There may be multiple transports with
766 different interface settings and the retry behaviour needs to be kept
769 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
771 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
773 JH/10 Bug 840: fix log_defer_output option of pipe transport
775 JH/11 Bug 830: use same host for all RCPTS of a message, even under
776 hosts_randomize. This matters a lot when combined with mua_wrapper.
778 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
779 ${quote_pgsql:<string>} operator.
781 JH/13 Bug 1708: avoid misaligned access in cached lookup.
783 JH/14 Change header file name for freeradius-client. Relevant if compiling
784 with Radius support; from the Gentoo tree and checked under Fedora.
786 JH/15 Bug 1712: Introduce $prdr_requested flag variable
788 JH/16 Bug 1714: Permit an empty string as expansion result for transport
789 option transport_filter, meaning no filtering.
791 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
793 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
794 defaults to "*" (all hosts). The variable is now available when not built
795 with TLS, default unset, mainly to enable keeping the testsuite sane.
796 If a server certificate is not supplied (via tls_certificate) an error is
797 logged, and clients will find TLS connections fail on startup. Presumably
798 they will retry in-clear.
799 Packagers of Exim are strongly encouraged to create a server certificate
800 at installation time.
802 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
803 with the $config_file variable.
805 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
806 in transport context, after the attempt, and per-recipient. The latter type
807 is per host attempted. The event data is the error message, and the errno
808 information encodes the lookup type (A vs. MX) used for the (first) host,
809 and the trailing two digits of the smtp 4xx response.
811 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
812 to write to mainlog (or rejectlog, paniclog) in the window between file
813 creation and permissions/ownership being changed. Particularly affects
814 installations where exicyclog is run as root, rather than exim user;
815 result is that the running daemon panics and dies.
817 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
819 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
820 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
821 "pri" and "weight". Note that the previous implicit priority given by the
822 list order is no longer honoured.
824 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
827 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
828 by defining SUPPORT_SOCKS.
830 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
831 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
832 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
833 variables are renamed to proxy_{local,external}_{address,port}.
835 JH/27 Move Internationalisation support from Experimental to mainline, enabled
836 for a build by defining SUPPORT_I18N
838 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
839 of the query string, and make ${quote_redis:} do that quoting.
841 JH/29 Move Events support from Experimental to mainline, enabled by default
842 and removable for a build by defining DISABLE_EVENT.
844 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
846 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
847 cached by the daemon.
849 JH/32 Move Redis support from Experimental to mainline, enabled for a build
850 by defining LOOKUP_REDIS. The libhiredis library is required.
852 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
853 keys are given for lookup.
855 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
856 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
857 only supported when built with TLS support. The PolarSSL SHA routines
858 are still used when the TLS library is too old for convenient support.
860 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
861 openssl_options), for security. OpenSSL forces this from version 1.1.0
862 server-side so match that on older versions.
864 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
865 allocation for $value could be released as the expansion processing
866 concluded, but leaving the global pointer active for it.
868 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
869 and to use the domains and local_parts ACL conditions.
871 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
872 incorrectly not doubled on cutthrough transmission, hence seen as a
873 body-termination at the receiving system - resulting in truncated mails.
874 Commonly the sender saw a TCP-level error, and retransmitted the message
875 via the normal store-and-forward channel. This could result in duplicates
876 received - but deduplicating mailstores were liable to retain only the
877 initial truncated version.
879 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
881 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
883 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
884 we're in there, support oversigning also; bug 1309.
886 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
888 HS/04 Add support for keep_environment and add_environment options.
890 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
891 either intentional arithmetic overflow during PRNG, or testing config-
894 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
895 delivery resulted in actual delivery. Cancel cutthrough before DATA
898 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
899 rejects a recipient: pass the reject to the originator.
901 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
902 Many were false-positives and ignorable, but it's worth fixing the
905 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
906 for the new environment-manipulation done at startup. Move the routines
907 from being local to tls.c to being global via the os.c file.
909 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
910 an extract embedded as result-arg for a map, the first arg for extract
911 is unavailable so we cannot tell if this is a numbered or keyed
912 extraction. Accept either.
918 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
921 JH/02 The smtp transport option "multi_domain" is now expanded.
923 JH/03 The smtp transport now requests PRDR by default, if the server offers
926 JH/04 Certificate name checking on server certificates, when exim is a client,
927 is now done by default. The transport option tls_verify_cert_hostnames
928 can be used to disable this per-host. The build option
929 EXPERIMENTAL_CERTNAMES is withdrawn.
931 JH/05 The value of the tls_verify_certificates smtp transport and main options
932 default to the word "system" to access the system default CA bundle.
933 For GnuTLS, only version 3.0.20 or later.
935 JH/06 Verification of the server certificate for a TLS connection is now tried
936 (but not required) by default. The verification status is now logged by
937 default, for both outbound TLS and client-certificate supplying inbound
940 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
943 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
944 Status Notification (bounce) messages are now MIME format per RFC 3464.
945 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
946 under the control of the dsn_advertise_hosts option, and routers may
947 have a dsn_lasthop option.
949 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
950 default, modifiable by a malware= option. The list separator for
951 the options can now be changed in the usual way. Bug 68.
953 JH/10 The smtp_receive_timeout main option is now expanded before use.
955 JH/11 The incoming_interface log option now also enables logging of the
956 local interface on delivery outgoing connections.
958 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
959 if the interface and destination host and port all match.
961 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
964 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
965 Patch from Andrew Lewis.
967 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
968 now supports optional time-restrictions, weighting, and priority
969 modifiers per server. Patch originally by <rommer@active.by>.
971 JH/16 The spamd_address main option now supports a mixed list of local
972 and remote servers. Remote servers can be IPv6 addresses, and
973 specify a port-range.
975 JH/17 Bug 68: The spamd_address main option now supports an optional
976 timeout value per server.
978 JH/18 Bug 1581: Router and transport options headers_add/remove can
979 now have the list separator specified.
981 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
984 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
987 JH/21 Support for the A6 type of dns record is withdrawn.
989 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
990 rather than the verbs used.
992 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
993 from 255 to 1024 chars.
995 JH/24 Verification callouts now attempt to use TLS by default.
997 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
998 are generic router options now. The defaults didn't change.
1000 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1001 Original patch from Alexander Shikoff, worked over by JH.
1003 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1004 files if ZCAT_COMMAND is not executable.
1006 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1008 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1010 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1011 Normally benign, it bites when the pair was led to by a CNAME;
1012 modern usage is to not canonicalize the domain to a CNAME target
1013 (and we were inconsistent anyway for A-only vs AAAA+A).
1015 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1017 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1018 when evaluating $sender_host_dnssec.
1020 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1021 $sender_helo_dnssec variable.
1023 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1025 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1027 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1029 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1030 documented as working, but never had. Support all but $spam_report.
1032 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1033 added for tls authenticator.
1035 HS/03 Add perl_taintmode main config option
1041 TL/01 When running the test suite, the README says that variables such as
1042 no_msglog_check are global and can be placed anywhere in a specific
1043 test's script, however it was observed that placement needed to be near
1044 the beginning for it to behave that way. Changed the runtest perl
1045 script to read through the entire script once to detect and set these
1046 variables, reset to the beginning of the script, and then run through
1047 the script parsing/test process like normal.
1049 TL/02 The BSD's have an arc4random API. One of the functions to induce
1050 adding randomness was arc4random_stir(), but it has been removed in
1051 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1052 function when detected.
1054 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1055 cause callback expansion.
1057 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1058 syntax errors in an expansion can be treated as a string instead of
1059 logging or causing an error, due to the internal use of bool_lax
1060 instead of bool when processing it.
1062 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1063 server certificates when making smtp deliveries.
1065 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1067 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1069 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1071 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1072 Merged patch from Sebastian Wiedenroth.
1074 JH/05 Fix results-pipe from transport process. Several recipients, combined
1075 with certificate use, exposed issues where response data items split
1076 over buffer boundaries were not parsed properly. This eventually
1077 resulted in duplicates being sent. This issue only became common enough
1078 to notice due to the introduction of connection certificate information,
1079 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1081 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1082 size buffer was used, resulting in syntax errors when an expansion
1085 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1086 version 3.3.6 or later.
1088 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1089 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1090 both become "event_action", the variables become $event_name, $event_data
1091 and $event_defer_errno. There is a new variable $verify_mode, usable in
1092 routers, transports and related events. The tls:cert event is now also
1093 raised for inbound connections, if the main configuration event_action
1096 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1097 early OCSP support, but no stapling (appears to be less than 1.0.0).
1099 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1100 server certificate names available under the smtp transport option
1101 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1104 JH/10 Time-related extraction expansions from certificates now use the main
1105 option "timezone" setting for output formatting, and are consistent
1106 between OpenSSL and GnuTLS compilations. Bug 1541.
1108 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1109 encoded parameter in the incoming message. Bug 1558.
1111 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1112 include certificate info, eximon was claiming there were spoolfile
1115 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1117 JH/14 Log delivery-related information more consistently, using the sequence
1118 "H=<name> [<ip>]" wherever possible.
1120 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1121 are problematic for Debian distribution, omit them from the release
1124 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1126 JH/16 Fix string representation of time values on 64bit time_t architectures.
1129 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1130 output list separator was used.
1135 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1136 checkers that were complaining about end of non-void function with no
1139 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1140 This was a regression introduced in 4.83 by another bugfix.
1142 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1144 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1145 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1151 TF/01 Correctly close the server side of TLS when forking for delivery.
1153 When a message was received over SMTP with TLS, Exim failed to clear up
1154 the incoming connection properly after forking off the child process to
1155 deliver the message. In some situations the subsequent outgoing
1156 delivery connection happened to have the same fd number as the incoming
1157 connection previously had. Exim would try to use TLS and fail, logging
1158 a "Bad file descriptor" error.
1160 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1161 utilities have not been installed.
1163 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1164 temporary space as the ACL may create new global variables.
1166 TL/01 LDAP support uses per connection or global context settings, depending
1167 upon the detected version of the libraries at build time.
1169 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1170 to extract and use the src ip:port in logging and expansions as if it
1171 were a direct connection from the outside internet. PPv2 support was
1172 updated based on HAProxy spec change in May 2014.
1174 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1176 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1177 Properly escape header and check for NULL return.
1179 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1182 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1184 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1185 characters in header names, implemented as a verify condition.
1186 Contributed by Michael Fischer v. Mollard.
1188 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1189 results permerror and temperror. Previous values are deprecated but
1190 still accepted. In a future release, err_perm and err_temp will be
1191 completely removed, which will be a backward incompatibility if the
1192 ACL tests for either of these two old results. Patch contributed by
1193 user bes-internal on the mailing list.
1195 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1197 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1198 selectors, in both main and reject logs.
1200 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1203 JH/07 Add malware type "sock" for talking to simple daemon.
1205 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1207 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1208 routers/transports under cutthrough routing.
1210 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1211 numbers. Touch up "bool" conditional to keep the same definition.
1213 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1215 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1217 JH/12 Expand items in router/transport headers_add or headers_remove lists
1218 individually rather than the list as a whole. Bug 1452.
1220 Required for reasonable handling of multiple headers_ options when
1221 they may be empty; requires that headers_remove items with embedded
1222 colons must have them doubled (or the list-separator changed).
1224 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1225 view the policy declared in the DMARC record. Currently, $dmarc_status
1226 is a combined value of both the record presence and the result of the
1229 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1231 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1232 dnslookup router and the smtp transport (applying to the forward
1235 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1236 of ldap servers used for a specific lookup. Patch provided by Heiko
1239 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1240 New variable $lookup_dnssec_authenticated for observability.
1242 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1243 Patch submitted by Lars Timman.
1245 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1247 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1248 Requires trusted mode and valid format message id, aborts otherwise.
1249 Patch contributed by Heiko Schlichting.
1251 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1252 certextract with support for various fields. Bug 1358.
1254 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1255 is requested by default, modifiable by smtp transport option
1258 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1259 operate on certificate variables to give certificate fingerprints
1260 Also new ${sha256:cert_variable}.
1262 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1264 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1265 Christian Aistleitner.
1267 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1269 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1270 file. Patch from Wolfgang Breyha.
1272 JH/25 Expand the coverage of the delivery $host and $host_address to
1273 client authenticators run in verify callout. Bug 1476.
1275 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1276 align with daemon_smtp_ports. Bug 72.
1278 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1279 support and error reporting did not work properly.
1281 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1282 and is readable. Patch from Andrew Colin Kissa.
1284 TL/14 Enhance documentation of ${run expansion and how it parses the
1285 commandline after expansion, particularly in the case when an
1286 unquoted variable expansion results in an empty value.
1288 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1290 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1291 by renaming away from it. Addresses GH issue 15, reported by
1294 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1296 TL/15 SECURITY: prevent double expansion in math comparison functions
1297 (can expand unsanitized data). Not remotely exploitable.
1304 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1306 PP/02 Make -n do something, by making it not do something.
1307 When combined with -bP, the name of an option is not output.
1309 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1312 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1313 $sender_host_name and config options to manage this, and basic check
1316 PP/05 DSCP support for outbound connections and control modifier for inbound.
1318 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1319 (Only plugin which currently uses this is kerberos4, which nobody should
1320 be using, but we should make it available and other future plugins might
1321 conceivably use it, even though it would break NAT; stuff *should* be
1322 using channel bindings instead).
1324 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1325 name; added for Sendmail compatibility; requires admin caller.
1326 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1327 just ignore it); requires trusted caller.
1328 Also parse but ignore: -Ac -Am -X<logfile>
1331 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1333 TL/02 Add +smtp_confirmation as a default logging option.
1335 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1336 Patch by Magnus Holmgren from 2007-02-20.
1338 TL/04 Bugzilla 1281 - Spec typo.
1339 Bugzilla 1283 - Spec typo.
1340 Bugzilla 1290 - Spec grammar fixes.
1342 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1344 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1346 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1347 mailing list by by Dmitry Isaikin.
1349 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1351 JH/02 Support "G" suffix to numbers in ${if comparisons.
1353 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1355 NM/01 Bugzilla 1197 - Spec typo
1356 Bugzilla 1196 - Spec examples corrections
1358 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1360 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1361 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1364 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1365 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1367 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1368 "acl {{name}{arg}...}", and optional args on acl condition
1371 JH/05 Permit multiple router/transport headers_add/remove lines.
1373 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1375 JH/07 Avoid using a waiting database for a single-message-only transport.
1376 Performance patch from Paul Fisher. Bugzilla 1262.
1378 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1381 JH/09 Add $headers_added variable, with content from use of ACL modifier
1382 add_header (but not yet added to the message). Bugzilla 199.
1384 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1385 Pulled from Bugzilla 817 by Wolfgang Breyha.
1387 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1389 (nb: this is the same fix as in Exim 4.80.1)
1391 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1394 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1397 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1399 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1400 advertise SMTP AUTH mechanism to us, instead of a generic
1401 protocol violation error. Also, make Exim more robust to bad
1402 data from the Dovecot auth socket.
1404 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1406 When a queue runner is handling a message, Exim first routes the
1407 recipient addresses, during which it prunes them based on the retry
1408 hints database. After that it attempts to deliver the message to
1409 any remaining recipients. It then updates the hints database using
1412 So if a recipient address works intermittently, it can get repeatedly
1413 deferred at routing time. The retry hints record remains fresh so the
1414 address never reaches the final cutoff time.
1416 This is a fairly common occurrence when a user is bumping up against
1417 their storage quota. Exim had some logic in its local delivery code
1418 to deal with this. However it did not apply to per-recipient defers
1419 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1421 This change adds a proper retry rule check during routing so that the
1422 final cutoff time is checked against the message's age. We only do
1423 this check if there is an address retry record and there is not a
1424 domain retry record; this implies that previous attempts to handle
1425 the address had the retry_use_local_parts option turned on. We use
1426 this as an approximation for the destination being like a local
1427 delivery, as in LMTP.
1429 I suspect this new check makes the old local delivery cutoff check
1430 redundant, but I have not verified this so I left the code in place.
1432 TF/02 Correct gecos expansion when From: is a prefix of the username.
1434 Test 0254 submits a message to Exim with the header
1438 When I ran the test suite under the user fanf2, Exim expanded
1439 the header to contain my full name, whereas it should have added
1440 a Resent-Sender: header. It erroneously treats any prefix of the
1441 username as equal to the username.
1443 This change corrects that bug.
1445 GF/01 DCC debug and logging tidyup
1446 Error conditions log to paniclog rather than rejectlog.
1447 Debug lines prefixed by "DCC: " to remove any ambiguity.
1449 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1451 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1452 Bug spotted by Jeremy Harris; was flawed since initial commit.
1453 Would have resulted in OCSP responses post-SNI triggering an Exim
1454 NULL dereference and crash.
1456 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1458 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1459 Bug detection, analysis and fix by Samuel Thibault.
1460 Bugzilla 1331, Debian bug #698092.
1462 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1464 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1465 Server implementation by Todd Lyons, client by JH.
1466 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1467 config variable "prdr_enable" controls whether the server
1468 advertises the facility. If the client requests PRDR a new
1469 acl_data_smtp_prdr ACL is called once for each recipient, after
1470 the body content is received and before the acl_smtp_data ACL.
1471 The client is controlled by both of: a hosts_try_prdr option
1472 on the smtp transport, and the server advertisement.
1473 Default client logging of deliveries and rejections involving
1474 PRDR are flagged with the string "PRDR".
1476 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1477 fclose(). Diagnosis by Todd Lyons.
1479 PP/17 Update configure.default to handle IPv6 localhost better.
1480 Patch by Alain Williams (plus minor tweaks).
1483 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1484 This is now consistent with GnuTLS, and is now documented: the
1485 previous undocumented portable approach to treating the option as
1486 unset was to force an expansion failure. That still works, and
1487 an empty string is now equivalent.
1489 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1490 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1491 not performing validation itself.
1493 PP/20 Added force_command boolean option to pipe transport.
1494 Patch from Nick Koston, of cPanel Inc.
1496 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1499 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1501 PP/21 Fix eximon continuous updating with timestamped log-files.
1502 Broken in a format-string cleanup in 4.80, missed when I repaired the
1503 other false fix of the same issue.
1504 Report and fix from Heiko Schlichting.
1507 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1508 Report from Prashanth Katuri.
1510 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1511 It's SecureTransport, so affects any MacOS clients which use the
1512 system-integrated TLS libraries, including email clients.
1514 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1515 using a MIME ACL for non-SMTP local injection.
1516 Report and assistance in diagnosis by Warren Baker.
1518 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1520 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1522 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1523 last id that failed so it may be referenced in subsequent ACL's.
1525 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1528 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1529 ldap library initialization, allowing self-signed CA's to be
1530 used. Also properly sets require_cert option later in code by
1531 using NULL (global ldap config) instead of ldap handle (per
1532 session). Bug diagnosis and testing by alxgomz.
1534 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1535 the src/util/ subdirectory.
1537 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1538 renamed to Transport Post Delivery Action by Jeremy Harris, as
1541 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1542 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1543 redis_servers = needs to be configured which will be used by the redis
1544 lookup. Patch from Warren Baker, of The Packet Hub.
1546 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1548 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1549 hostname or reverse DNS when processing a host list. Used suggestions
1550 from multiple comments on this bug.
1552 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1554 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1555 Missed a few lines, added it to make the runtest require no keyboard
1558 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1559 contains upper case chars. Make router use caseful_local_part.
1561 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1562 support when GnuTLS has been built with p11-kit.
1568 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1570 This, or similar/improved, will also be change PP/11 of 4.82.
1576 PP/01 Handle short writes when writing local log-files.
1577 In practice, only affects FreeBSD (8 onwards).
1578 Bugzilla 1053, with thanks to Dmitry Isaikin.
1580 NM/01 Bugzilla 949 - Documentation tweak
1582 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1585 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1587 PP/02 Implemented gsasl authenticator.
1589 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1591 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1592 `pkg-config foo` for cflags/libs.
1594 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1595 with rest of GSASL and with heimdal_gssapi.
1597 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1598 `pkg-config foo` for cflags/libs for the TLS implementation.
1600 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1601 properties get this fed in as external SSF. A number of robustness
1602 and debugging improvements to the cyrus_sasl authenticator.
1604 PP/08 cyrus_sasl server now expands the server_realm option.
1606 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1607 Patch by Jeremy Harris.
1609 PP/10 Added dbmjz lookup type.
1611 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1613 PP/12 MAIL args handles TAB as well as SP, for better interop with
1614 non-compliant senders.
1615 Analysis and variant patch by Todd Lyons.
1617 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1618 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1619 Patch from Dirk Mueller <dmueller@suse.com>
1621 PP/13 tls_peerdn now print-escaped for spool files.
1622 Observed some $tls_peerdn in wild which contained \n, which resulted
1623 in spool file corruption.
1625 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1626 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1627 or write after TLS renegotiation, which otherwise led to messages
1630 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1631 as a tracking header (ie: a signed header comes before the signature).
1632 Patch from Wolfgang Breyha.
1634 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1635 comma-sep list; embedded commas doubled.
1637 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1639 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1641 Report and patch from Dmitry Banschikov.
1643 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1644 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1645 failures. We appear to now support TLS1.1+ with Exim.
1647 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1648 lets Exim select keys and certificates based upon TLS SNI from client.
1649 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1650 before an outbound SMTP session. New log_selector, +tls_sni.
1652 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1653 NULL dereference. Report and patch from Alun Jones.
1655 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1656 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1657 Not seeing resolver debug output on NetBSD, but suspect this is a
1658 resolver implementation change.
1660 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1661 Left warnings. Added "eximon gdb" invocation mode.
1663 PP/21 Defaulting "accept_8bitmime" to true, not false.
1665 PP/22 Added -bw for inetd wait mode support.
1667 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1668 locate the relevant includes and libraries. Made this the default.
1670 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1671 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1673 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1674 This may cause build issues on older platforms.
1676 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1677 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1678 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1679 Added SNI support via GnuTLS too.
1680 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1682 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1684 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1685 Applied second patch from Janne, implementing suggestion to default
1686 multiple-strings-in-record handling to match SPF spec.
1688 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1690 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1691 read-only, out of scope).
1692 Patch from Wolfgang Breyha, report from Stuart Northfield.
1694 PP/29 Fix three issues highlighted by clang analyser static analysis.
1695 Only crash-plausible issue would require the Cambridge-specific
1696 iplookup router and a misconfiguration.
1697 Report from Marcin Mirosław.
1699 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1701 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1702 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1703 As part of this, removing so much warning spew let me fix some minor
1704 real issues in debug logging.
1706 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1707 assignment on my part. Fixed.
1709 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1710 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1711 Janne Snabb (who went above and beyond: thank you).
1713 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1714 string otherwise requires a connection and a bunch more work and it's
1715 relatively easy to get wrong. Should also expose TLS library linkage
1718 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1719 64-bit ${eval} (JH/03).
1721 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1722 GNU libc to support some of the 64-bit stuff, should not lead to
1723 conflicts. Defined before os.h is pulled in, so if a given platform
1724 needs to override this, it can.
1726 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1727 protection layer was required, which is not implemented.
1728 Bugzilla 1254, patch from Wolfgang Breyha.
1730 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1731 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1732 tls_dhparam take prime identifiers. Also unbreak combination of
1733 OpenSSL+DH_params+TLSSNI.
1735 PP/39 Disable SSLv2 by default in OpenSSL support.
1741 PP/01 Solaris build fix for Oracle's LDAP libraries.
1742 Bugzilla 1109, patch from Stephen Usher.
1744 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1746 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1749 TF/02 Fix a couple more cases where we did not log the error message
1750 when unlink() failed. See also change 4.74-TF/03.
1752 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1753 lock up or crash if it happened to be inside a call to libc when it
1754 got a SIGUSR1 from exiwhat.
1756 The SIGUSR1 handler appends the current process status to the process
1757 log which is later printed by exiwhat. It used to use the general
1758 purpose logging code to do this, but several functions it calls are
1759 not safe for signals.
1761 The new output code in the SIGUSR1 handler is specific to the process
1762 log, and simple enough that it's easy to inspect for signal safety.
1763 Removing some special cases also simplifies the general logging code.
1764 Removing the spurious timestamps from the process log simplifies
1767 TF/04 Improved ratelimit ACL condition.
1769 The /noupdate option has been deprecated in favour of /readonly which
1770 has clearer semantics. The /leaky, /strict, and /readonly update modes
1771 are mutually exclusive. The update mode is no longer included in the
1772 database key; it just determines when the database is updated. (This
1773 means that when you upgrade Exim will forget old rate measurements.)
1775 Exim now checks that the per_* options are used with an update mode that
1776 makes sense for the current ACL. For example, when Exim is processing a
1777 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1778 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1779 must specify per_mail/readonly. If you omit the update mode it defaults to
1780 /leaky where that makes sense (as before) or /readonly where required.
1782 The /noupdate option is now undocumented but still supported for
1783 backwards compatibility. It is equivalent to /readonly except that in
1784 ACLs where /readonly is required you may specify /leaky/noupdate or
1785 /strict/noupdate which are treated the same as /readonly.
1787 A useful new feature is the /count= option. This is a generalization
1788 of the per_byte option, so that you can measure the throughput of other
1789 aggregate values. For example, the per_byte option is now equivalent
1790 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1792 The per_rcpt option has been generalized using the /count= mechanism
1793 (though it's more complicated than the per_byte equivalence). When it is
1794 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1795 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1796 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1797 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1798 Note that using per_rcpt with a non-readonly update mode in more than
1799 one ACL will cause the recipients to be double-counted. (The per_mail
1800 and per_byte options don't have this problem.)
1802 The handling of very low rates has changed slightly. If the computed rate
1803 is less than the event's count (usually one) then this event is the first
1804 after a long gap. In this case the rate is set to the same as this event's
1805 count, so that the first message of a spam run is counted properly.
1807 The major new feature is a mechanism for counting the rate of unique
1808 events. The new per_addr option counts the number of different
1809 recipients that someone has sent messages to in the last time period. It
1810 behaves like per_rcpt if all the recipient addresses are different, but
1811 duplicate recipient addresses do not increase the measured rate. Like
1812 the /count= option this is a general mechanism, so the per_addr option
1813 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1814 example, measure the rate that a client uses different sender addresses
1815 with the options per_mail/unique=$sender_address. There are further
1816 details in the main documentation.
1818 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1820 TF/06 Removed a few PCRE remnants.
1822 TF/07 Automatically extract Exim's version number from tags in the git
1823 repository when doing development or release builds.
1825 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1826 Bugzilla 879. Patch from Paul Fisher.
1828 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1829 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1832 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1834 PP/05 Variable $av_failed, true if the AV scanner deferred.
1835 Bugzilla 1078. Patch from John Horne.
1837 PP/06 Stop make process more reliably on build failure.
1838 Bugzilla 1087. Patch from Heiko Schlittermann.
1840 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1841 Bugzilla 1089. Patch from Heiko Schlittermann.
1843 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1844 Bugzilla 1131. Patch from Holger Weiß.
1846 PP/09 Handle IPv6 addresses with SPF.
1847 Bugzilla 860. Patch from Wolfgang Breyha.
1849 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1851 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1854 PP/11 match_* no longer expand right-hand-side by default.
1855 New compile-time build option, EXPAND_LISTMATCH_RHS.
1856 New expansion conditions, "inlist", "inlisti".
1858 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1860 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1862 PP/14 fix log_write() format string regression from TF/03.
1863 Bugzilla 1152. Patch from Dmitry Isaikin.
1869 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1871 PP/02 Harmonised TLS library version reporting; only show if debugging.
1872 Layout now matches that introduced for other libraries in 4.74 PP/03.
1874 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1876 PP/04 New "dns_use_edns0" global option.
1878 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1881 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1882 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1884 TK/01 Updated PolarSSL code to 0.14.2.
1885 Bugzilla 1097. Patch from Andreas Metzler.
1887 PP/07 Catch divide-by-zero in ${eval:...}.
1888 Fixes bugzilla 1102.
1890 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1893 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1894 format-string attack -- SECURITY: remote arbitrary code execution.
1896 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1897 time unintentionally subject to list matching rules, letting the header
1898 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1899 arbitrary string expansion). This allowed for information disclosure.
1901 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1902 INT_MIN/-1 -- value coerced to INT_MAX.
1908 NM/01 Workaround for PCRE version dependency in version reporting
1911 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1912 This fixes portability to compilers other than gcc, notably
1913 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1915 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1916 makefiles for portability to HP-UX and POSIX correctness.
1918 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1919 Also via indented variable definition in the Makefile.
1920 (Debugging by Oliver Heesakkers).
1922 PP/02 Restore caching of spamd results with expanded spamd_address.
1923 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1925 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1926 Improves build reliability. Fix from: Frank Elsner
1928 NM/02 Fix wide character breakage in the rfc2047 coding
1929 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1931 NM/03 Allow underscore in dnslist lookups
1932 Fixes bug 1026. Patch from Graeme Fowler
1934 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1935 Code patches from Adam Ciarcinski of NetBSD.
1937 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1940 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1941 is logged, to avoid truncation. Patch from John Horne.
1943 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1944 Patch from Jakob Hirsch.
1946 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1947 SQL string expansion failure details.
1948 Patch from Andrey Oktyabrski.
1950 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1951 Patch from Simon Arlott.
1953 PP/09 New lookups functionality failed to compile on old gcc which rejects
1954 extern declarations in function scope.
1955 Patch from Oliver Fleischmann
1957 PP/10 Use sig_atomic_t for flags set from signal handlers.
1958 Check getgroups() return and improve debugging.
1959 Fixed developed for diagnosis in bug 927 (which turned out to be
1962 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1963 Patch from Mark Zealey.
1965 PP/12 Bugzilla 1056: Improved spamd server selection.
1966 Patch from Mark Zealey.
1968 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1969 Based on patch from Heiko Schlittermann.
1971 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1972 Patch from Uwe Doering, sign-off by Michael Haardt.
1974 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1975 variable declaration deep within a block. Bug and patch from
1978 PP/15 lookups-Makefile IRIX compatibility coercion.
1980 PP/16 Make DISABLE_DKIM build knob functional.
1982 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1983 Patch by Simon Arlott
1985 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1986 variable argument macros. Our copy now differs from upstream.
1992 TF/01 Failure to get a lock on a hints database can have serious
1993 consequences so log it to the panic log.
1995 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1996 controlled using the smtp_confirmation log selector.
1998 TF/03 Include the error message when we fail to unlink a spool file.
2000 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2001 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2002 for maintaining out-of-tree patches for some time.
2004 PP/01 Bugzilla 139: Documentation and portability issues.
2005 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2006 Handle per-OS dynamic-module compilation flags.
2008 PP/02 Let /dev/null have normal permissions.
2009 The 4.73 fixes were a little too stringent and complained about the
2010 permissions on /dev/null. Exempt it from some checks.
2011 Reported by Andreas M. Kirchwitz.
2013 PP/03 Report version information for many libraries, including
2014 Exim version information for dynamically loaded libraries. Created
2015 version.h, now support a version extension string for distributors
2016 who patch heavily. Dynamic module ABI change.
2018 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2019 privilege escalation vulnerability whereby the Exim run-time user
2020 can cause root to append content of the attacker's choosing to
2023 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2026 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2027 If dropping privileges for untrusted macros, we disabled normal logging
2028 on the basis that it would fail; for the Exim run-time user, this is not
2029 the case, and it resulted in successful deliveries going unlogged.
2030 Fixed. Reported by Andreas Metzler.
2036 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2037 only prepend for the Resent-* case. Fixes regression introduced in
2038 Exim 4.70 by NM/22 for Bugzilla 607.
2040 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2041 increasing numbers of administrators be bitten by this.
2043 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2045 PP/03 Bugzilla 994: added openssl_options main configuration option.
2047 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2049 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2051 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2053 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2054 without a peer certificate, leading to a segfault because of an
2055 assumption that peers always have certificates. Be a little more
2056 paranoid. Problem reported by Martin Tscholak.
2058 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2059 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2060 NB: ClamAV planning to remove STREAM in "middle of 2010".
2061 CL also introduces -bmalware, various -d+acl logging additions and
2062 more caution in buffer sizes.
2064 PP/09 Implemented reverse_ip expansion operator.
2066 PP/10 Bugzilla 937: provide a "debug" ACL control.
2068 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2070 PP/12 Bugzilla 973: Implement --version.
2072 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2074 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2076 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2078 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2079 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2080 ignore trailing whitespace.
2082 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2084 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2087 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2088 Notification from Dr Andrew Aitchison.
2090 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2091 ExtendedDetectionInfo response format.
2092 Notification from John Horne.
2094 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2097 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2098 XSL and documented dependency on system catalogs, with examples of how
2101 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2104 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2105 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2106 configuration file which is writeable by the Exim user or group.
2108 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2109 of configuration files to cover files specified with the -C option if
2110 they are going to be used with root privileges, not just the default
2113 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2114 option (effectively making it always true).
2116 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2117 files to be used while preserving root privileges.
2119 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2120 that rogue child processes cannot use them.
2122 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2123 run-time user, instead of root.
2125 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2126 Exim run-time user without dropping privileges.
2128 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2129 result string, instead of calling string_vformat() twice with the same
2132 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2133 for other users. Others should always drop root privileges if they use
2134 -C on the command line, even for a whitelisted configure file.
2136 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2138 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2144 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2145 $data_path, and $header_path variables; fixed documentation bugs and
2148 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2149 exipick to access non-standard spools, including the "frozen" queue
2152 NM/01 Bugzilla 965: Support mysql stored procedures.
2153 Patch from Alain Williams
2155 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2157 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2158 Patch from Andreas Metzler
2160 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2161 Patch from Kirill Miazine
2163 NM/05 Bugzilla 671: Added umask to procmail example.
2165 JJ/03 installed exipick 20100323.0, fixing doc bug
2167 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2168 directory. Notification and patch from Dan Rosenberg.
2170 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2172 TK/02 Improve log output when DKIM signing operation fails.
2174 MH/01 Treat the transport option dkim_domain as a colon separated
2175 list, not as a single string, and sign the message with each element,
2176 omitting multiple occurences of the same signer.
2178 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2179 Bugzilla 985, 986. Patch by Simon Arlott
2181 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2182 Patch by Simon Arlott
2184 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2185 MBX locking. Notification from Dan Rosenberg.
2191 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2193 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2195 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2197 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2199 NM/04 Bugzilla 915: Flag broken perl installation during build.
2205 TK/01 Added patch by Johannes Berg that expands the main option
2206 "spamd_address" if it starts with a dollar sign.
2208 TK/02 Write list of recipients to X-Envelope-Sender header when building
2209 the mbox-format spool file for content scanning (suggested by Jakob
2212 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2213 (http://www.dcc-servers.net/) support via dccifd. Activated by
2214 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2216 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2217 by Mark Daniel Reidel <mr@df.eu>.
2219 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2220 When building exim an external PCRE library is now needed -
2221 PCRE is a system library on the majority of modern systems.
2222 See entry on PCRE_LIBS in EDITME file.
2224 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2225 conversation. Added nologin parameter to request.
2226 Patch contributed by Kirill Miazine.
2228 TF/01 Do not log submission mode rewrites if they do not change the address.
2230 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2232 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2233 log files in place. Contributed by Roberto Lima.
2235 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2237 TF/03 Bugzilla 615: When checking the local_parts router precondition
2238 after a local_part_suffix or local_part_prefix option, Exim now
2239 does not use the address's named list lookup cache, since this
2240 contains cached lookups for the whole local part.
2242 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2243 Robert Millan. Documentation is in experimental-spec.txt.
2245 TF/04 Bugzilla 668: Fix parallel build (make -j).
2247 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2249 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2250 Patch provided by Jan Srzednicki.
2252 TF/05 Leading white space used to be stripped from $spam_report which
2253 wrecked the formatting. Now it is preserved.
2255 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2256 that they are available at delivery time.
2258 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2260 TF/08 TLS error reporting now respects the incoming_interface and
2261 incoming_port log selectors.
2263 TF/09 Produce a more useful error message if an SMTP transport's hosts
2264 setting expands to an empty string.
2266 NM/06 Bugzilla 744: EXPN did not work under TLS.
2267 Patch provided by Phil Pennock.
2269 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2270 Patch provided by Richard Godbee.
2272 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2273 acl_smtp_notquit, added index entry.
2275 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2276 Patch provided by Eugene Bujak.
2278 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2279 accept(). Patch provided by Maxim Dounin.
2281 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2282 Patch provided by Phil Pennock.
2284 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2286 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2287 Patch provided by Brad "anomie" Jorsch.
2289 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2290 Patch provided by Dean Brooks.
2292 TK/05 Add native DKIM support (does not depend on external libraries).
2294 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2295 Patch provided by Graeme Fowler.
2297 NM/16 Bugzilla 851: Documentation example syntax fix.
2299 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2301 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2304 NM/19 Bugzilla 745: TLS version reporting.
2305 Patch provided by Phil Pennock.
2307 NM/20 Bugzilla 167: bool: condition support.
2308 Patch provided by Phil Pennock.
2310 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2311 clients. Patch provided by Phil Pennock.
2313 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2314 Patch provided by Brad "anomie" Jorsch.
2316 NM/23 Bugzilla 687: Fix misparses in eximstats.
2317 Patch provided by Heiko Schlittermann.
2319 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2320 Patch provided by Heiko Schlittermann.
2322 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2323 plus update to original patch.
2325 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2327 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2328 Patch provided by David Brownlee.
2330 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2332 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2334 NM/30 Bugzilla 888: TLS documentation bugfixes.
2336 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2338 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2339 Unlike the original bugzilla I have changed all shell scripts in src tree.
2341 NM/33 Bugzilla 898: Transport filter timeout fix.
2342 Patch by Todd Rinaldo.
2344 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2345 Patch by Serge Demonchaux.
2347 NM/35 Bugzilla 39: Base64 decode bug fixes.
2348 Patch by Jakob Hirsch.
2350 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2352 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2354 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2356 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2362 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2363 ALT-N's libdkim that I have put here:
2364 http://duncanthrax.net/exim-experimental/
2366 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2367 were called 'true' and it seems that C99 defines that as a reserved
2368 keyword to be used with 'bool' variable types. That means you could
2369 not include C99-style headers which use bools without triggering
2370 build errors in sieve.c.
2372 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2373 as mailq or other aliases. Changed the --help handling significantly
2374 to do whats expected. exim_usage() emits usage/help information.
2376 SC/01 Added the -bylocaldomain option to eximstats.
2378 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2380 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2382 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2388 PH/01 Another patch from the Sieve maintainer.
2390 PH/02 When an IPv6 address is converted to a string for single-key lookup
2391 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2392 dots are used instead of colons so that keys in lsearch files need not
2393 contain colons. This was done some time before quoting was made available
2394 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2395 (notated using the quote facility) so as to distinguish them from IPv4
2396 keys. This meant that lookups for IP addresses in host lists did not work
2397 for iplsearch lookups.
2399 This has been fixed by arranging for IPv6 addresses to be expressed with
2400 colons if the lookup type is iplsearch. This is not incompatible, because
2401 previously such lookups could never work.
2403 The situation is now rather anomalous, since one *can* have colons in
2404 ordinary lsearch keys. However, making the change in all cases is
2405 incompatible and would probably break a number of configurations.
2407 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2410 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2411 conversion specification without a maximum field width, thereby enabling
2412 a rogue spamd server to cause a buffer overflow. While nobody in their
2413 right mind would setup Exim to query an untrusted spamd server, an
2414 attacker that gains access to a server running spamd could potentially
2415 exploit this vulnerability to run arbitrary code as the Exim user.
2417 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2418 $primary_hostname instead of what libspf2 thinks the hosts name is.
2420 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2421 a directory entry by the name of the lookup key. Previously, if a
2422 symlink pointed to a non-existing file or a file in a directory that
2423 Exim lacked permissions to read, a lookup for a key matching that
2424 symlink would fail. Now it is enough that a matching directory entry
2425 exists, symlink or not. (Bugzilla 503.)
2427 PH/03 The body_linecount and body_zerocount variables are now exported in the
2430 PH/04 Added the $dnslist_matched variable.
2432 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2433 This means they are set thereafter only if the connection becomes
2436 PH/06 Added the client_condition to authenticators so that some can be skipped
2437 by clients under certain conditions.
2439 PH/07 The error message for a badly-placed control=no_multiline_responses left
2440 "_responses" off the end of the name.
2442 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2444 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2445 (without spaces) instead of just copying the configuration text.
2447 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2449 PH/11 Added $max_received_linelength.
2451 PH/12 Added +ignore_defer and +include_defer to host lists.
2453 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2454 way in which PCRE > 7.0 is built.
2456 PH/14 Implemented queue_only_load_latch.
2458 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2459 MAIL command. The effect was to mangle the value on 64-bit systems.
2461 PH/16 Another patch from the Sieve maintainer.
2463 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2465 PH/18 If a system quota error occurred while trying to create the file for
2466 a maildir delivery, the message "Mailbox is full" was not appended to the
2467 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2468 only to a quota excession during the actual writing of the file.
2470 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2471 characters?) which causes problems in log lines. The DN values are now
2472 passed through string_printing() before being added to log lines.
2474 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2475 and InterBase are left for another time.)
2477 PH/21 Added message_body_newlines option.
2479 PH/22 Guard against possible overflow in moan_check_errorcopy().
2481 PH/23 POSIX allows open() to be a macro; guard against that.
2483 PH/24 If the recipient of an error message contained an @ in the local part
2484 (suitably quoted, of course), incorrect values were put in $domain and
2485 $local_part during the evaluation of errors_copy.
2491 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2492 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2495 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2496 issue a MAIL command.
2498 PH/02 In an ACL statement such as
2500 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2502 if a client was not listed at all, or was listed with a value other than
2503 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2504 the condition was not true (as it should be), so access was not denied.
2505 The bug was that the ! inversion was incorrectly passed on to the second
2506 item. This has been fixed.
2508 PH/03 Added additional dnslists conditions == and =& which are different from
2509 = and & when the dns lookup returns more than one IP address.
2511 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2512 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2514 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2515 FSYNC, which compiles an option called disable_fsync that allows for
2516 bypassing fsync(). The documentation is heavily laced with warnings.
2518 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2520 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2521 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2522 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2523 including adding "make clean"; (3) Added -fPIC when compiling the test
2524 dynamically loaded module, to get rid of a warning.
2526 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2527 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2528 = 0s. The bug is otherwise harmless.
2530 PH/07 There was a bug in the dovecot authenticator such that the value of
2531 $auth1 could be overwritten, and so not correctly preserved, after a
2532 successful authentication. This usually meant that the value preserved by
2533 the server_setid option was incorrect.
2535 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2537 PH/09 Installed PCRE release 7.0.
2539 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2540 run for batched SMTP input. It is now run at the start of every message
2541 in the batch. While fixing this I discovered that the process information
2542 (output by running exiwhat) was not always getting set for -bs and -bS
2543 input. This is fixed, and it now also says "batched" for BSMTP.
2545 PH/11 Added control=no_pipelining.
2547 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2548 patch, slightly modified), and move the expansion of helo_data till after
2549 the connection is made in the smtp transport (so it can use these
2552 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2554 PH/14 Added log_selector = +pid.
2556 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2558 PH/16 Add ${if forany and ${if forall.
2560 PH/17 Added dsn_from option to vary the From: line in DSNs.
2562 PH/18 Flush SMTP output before performing a callout, unless control =
2563 no_callout_flush is set.
2565 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2566 was true (the default) a successful delivery failed to delete the retry
2567 item, thus causing premature timeout of the address. The bug is now
2570 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2572 PH/21 Long custom messages for fakedefer and fakereject are now split up
2573 into multiline responses in the same way that messages for "deny" and
2574 other ACL rejections are.
2576 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2577 with slight modification.
2579 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2580 draft, changing the syntax and factoring some duplicate code".
2582 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2583 for deliveries of the second and subsequent messages over the same SMTP
2586 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2587 ${reduce, with only minor "tidies".
2589 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2591 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2592 expansion side effects.
2594 PH/27 When a message times out after an over-quota error from an Exim-imposed
2595 quota, the bounce message says "mailbox is full". This message was not
2596 being given when it was a system quota that was exceeded. It now should
2599 MH/03 Made $recipients available in local_scan(). local_scan() already has
2600 better access to the recipient list through recipients_list[], but
2601 $recipients can be useful in postmaster-provided expansion strings.
2603 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2604 in the case of a MAIL command with additional options following the
2605 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2606 were accidentally chopped off.
2608 PH/29 SMTP synchronization checks are implemented when a command is read -
2609 there is a check that no more input is waiting when there shouldn't be
2610 any. However, for some commands, a delay in an ACL can mean that it is
2611 some time before the response is written. In this time, more input might
2612 arrive, invalidly. So now there are extra checks after an ACL has run for
2613 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2614 pipelining has not been advertised.
2616 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2618 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2619 returned in the SMTP response when smtp_return_error_details was false.
2620 This has been fixed.
2622 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2623 instead of the C I/O that was originally supplied, because problems were
2624 reported on Solaris.
2626 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2627 Exim which did not show up earlier: it was assuming that a call to
2628 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2629 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2630 was a macro that became an assignment, so it seemed to work. This has
2631 changed to a proper function call with a void return, hence the compile
2632 error. Exim's code has been fixed.
2634 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2637 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2639 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2641 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2642 because it turns out to be too big or there is a timeout) there is no
2643 "Completed" line in the log. When some messages of this type were
2644 selected by exigrep, they were listed as "not completed". Others were
2645 picked up by some special patterns. I have improved the selection
2646 criteria to be more general.
2648 PH/38 The host_find_failed option in the manualroute router can now be set
2649 to "ignore", to completely ignore a host whose IP address cannot be
2650 found. If all hosts are ignored, the behaviour is controlled by the new
2651 host_all_ignored option.
2653 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2654 homing or because of multiple MX records with /mx) generated more than
2655 one IP address, and the following item turned out to be the local host,
2656 all the secondary addresses of the first item were incorrectly removed
2657 from the list, along with the local host and any following hosts (which
2658 is what is supposed to happen).
2660 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2661 whoever called Exim into the -H file. In the case of the daemon it was
2662 behaving confusingly. When first started, it used values for whoever
2663 started the daemon, but after a SIGHUP it used the Exim user (because it
2664 calls itself on a restart). I have changed the code so that it now always
2667 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2668 message are rejected with the same error (e.g. no authentication or bad
2669 sender address), and a DATA command is nevertheless sent (as can happen
2670 with PIPELINING or a stupid MUA), the error message that was given to the
2671 RCPT commands is included in the rejection of the DATA command. This is
2672 intended to be helpful for MUAs that show only the final error to their
2675 PH/42 Another patch from the Sieve maintainer.
2677 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2678 Eximstats - Fixed some broken HTML links and added missing column headers
2680 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2681 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2683 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2685 PH/43 Yet another patch from the Sieve maintainer.
2687 PH/44 I found a way to check for a TCP/IP connection going away before sending
2688 the response to the final '.' that terminates a message, but only in the
2689 case where the client has not sent further data following the '.'
2690 (unfortunately, this is allowed). However, in many cases there won't be
2691 any further data because there won't be any more messages to send. A call
2692 to select() can be used: if it shows that the input is "ready", there is
2693 either input waiting, or the socket has been closed. An attempt to read
2694 the next input character can distinguish the two cases. Previously, Exim
2695 would have sent an OK response which the client would never have see.
2696 This could lead to message repetition. This fix should cure that, at
2697 least in a lot of common cases.
2699 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2700 advertised in response to EHLO.
2706 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2707 fixed by 4.65/MH/01 (is this a record?) are fixed:
2709 (i) An empty string was always treated as zero by the numeric comparison
2710 operators. This behaviour has been restored.
2712 (ii) It is documented that the numeric comparison operators always treat
2713 their arguments as decimal numbers. This was broken in that numbers
2714 starting with 0 were being interpreted as octal.
2716 While fixing these problems I realized that there was another issue that
2717 hadn't been noticed. Values of message_size_limit (both the global option
2718 and the transport option) were treated as octal if they started with 0.
2719 The documentation was vague. These values are now always treated as
2720 decimal, and I will make that clear in the documentation.
2726 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2727 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2730 MH/01 Don't check that the operands of numeric comparison operators are
2731 integers when their expansion is in "skipping" mode (fixes bug
2732 introduced by 4.64-PH/07).
2734 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2735 child addresses, Exim now panics and dies. Previously, because the count
2736 is held in a short int, deliveries were likely to be lost. As such a
2737 large number of recipients for a single message is ridiculous
2738 (performance will be very, very poor), I have chosen to impose a limit
2739 rather than extend the field.
2745 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2746 leftover -K file (the existence of which was triggered by #402).
2747 While we were at it, introduced process PID as part of the -K
2748 filename. This should rule out race conditions when creating
2751 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2752 processing considerably. Previous code took too long for large mails,
2753 triggering a timeout which in turn triggers #401.
2755 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2756 in the DK code in transports.c. sendfile() is not really portable,
2757 hence the _LINUX specificness.
2759 TF/01 In the add_headers option to the mail command in an Exim filter,
2760 there was a bug that Exim would claim a syntax error in any
2761 header after the first one which had an odd number of characters
2764 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2765 callout verification, Exim cached a "reject" for the entire domain. This
2766 is correct for most verifications, but it is not correct for a recipient
2767 verification with use_sender or use_postmaster set, because in that case
2768 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2769 case of MAIL FROM:<> rejection from other early rejections (e.g.
2770 rejection of HELO). When verifying a recipient using a non-null MAIL
2771 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2772 Whatever the result of the callout, the value of the domain cache is
2773 left unchanged (for any other kind of callout, getting as far as trying
2774 RCPT means that the domain itself is ok).
2776 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2779 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2780 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2781 session, the daemon ignores ECONNECT errors and logs others; it now
2782 ignores EPIPE as well.
2784 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2785 (quoted-printable decoding).
2787 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2788 later the small subsequent patch to fix an introduced bug.
2790 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2792 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2794 PH/08 An error is now given if message_size_limit is specified negative.
2796 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2797 to be given (somewhat) arbitrary names.
2799 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2802 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2803 miscellaneous code fixes
2805 PH/10 Added the log_reject_target ACL modifier to specify where to log
2808 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2809 hostname. This is wrong, because it relates to the incoming message (and
2810 probably the interface on which it is arriving) and not to the outgoing
2811 callout (which could be using a different interface). This has been
2812 changed to use the value of the helo_data option from the smtp transport
2813 instead - this is what is used when a message is actually being sent. If
2814 there is no remote transport (possible with a router that sets up host
2815 addresses), $smtp_active_hostname is used.
2817 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2818 tweaks were necessary in order to get it to work (see also 21 below):
2819 (a) The code assumed that strncpy() returns a negative number on buffer
2820 overflow, which isn't the case. Replaced with Exim's string_format()
2822 (b) There were several signed/unsigned issues. I just did the minimum
2823 hacking in of casts. There is scope for a larger refactoring.
2824 (c) The code used strcasecmp() which is not a standard C function.
2825 Replaced with Exim's strcmpic() function.
2826 (d) The code set only $1; it now sets $auth1 as well.
2827 (e) A simple test gave the error "authentication client didn't specify
2828 service in request". It would seem that Dovecot has changed its
2829 interface. Fortunately there's a specification; I followed it and
2830 changed what the client sends and it appears to be working now.
2832 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2835 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2836 address A is aliased to B and C, where B exists and C does not. Without
2837 -v the output is "A verified" because verification stops after a
2838 successful redirection if more than one address is generated. However,
2839 with -v the child addresses are also verified. Exim was outputting "A
2840 failed to verify" and then showing the successful verification for C,
2841 with its parentage. It now outputs "B failed to verify", showing B's
2842 parentage before showing the successful verification of C.
2844 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2845 look up a TXT record in a specific list after matching in a combined
2848 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2849 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2850 they consult the DNS. I had assumed they would set it the way they
2851 wanted; and indeed my experiments on Linux seem to show that in some
2852 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2853 To be on the safe side, however, I have now made the interface to
2854 host_find_byname() similar to host_find_bydns(), with an argument
2855 containing the DNS resolver options. The host_find_byname() function now
2856 sets these options at its start, just as host_find_bydns() does. The smtp
2857 transport options dns_qualify_single and dns_search_parents are passed to
2858 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2859 of host_find_byname() use the default settings of RES_DEFNAMES
2860 (qualify_single) but not RES_DNSRCH (search_parents).
2862 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2863 spool_read_header() do less string testing, by means of a preliminary
2864 switch on the second character of optional "-foo" lines. (This is
2865 overdue, caused by the large number of possibilities that now exist.
2866 Originally there were few.) While I was there, I also converted the
2867 str(n)cmp tests so they don't re-test the leading "-" and the first
2868 character, in the hope this might squeeze out yet more improvement.
2870 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2871 flag allowing group syntax was set by the header_syntax check but not
2872 turned off, possible causing trouble later; (2) The flag was not being
2873 set at all for the header_verify test, causing "group"-style headers to
2874 be rejected. I have now set it in this case, and also caused header_
2875 verify to ignore an empty address taken from a group. While doing this, I
2876 came across some other cases where the code for allowing group syntax
2877 while scanning a header line wasn't quite right (mostly, not resetting
2878 the flag correctly in the right place). These bugs could have caused
2879 trouble for malformed header lines. I hope it is now all correct.
2881 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2882 with the "reply" argument non-NULL. The code, however (which originally
2883 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2884 but it didn't always do it. This confused somebody who was copying the
2885 code for some other use. I have removed all the tests.
2887 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2888 feature that was used to support insecure browsers during the U.S. crypto
2889 embargo. It requires special client support, and Exim is probably the
2890 only MTA that supported it -- and would never use it because real RSA is
2891 always available. This code has been removed, because it had the bad
2892 effect of slowing Exim down by computing (never used) parameters for the
2893 RSA_EXPORT functionality.
2895 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2896 authenticator to fail if there's a tab character in the incoming data
2897 (there should never be unless someone is messing about, as it's supposed
2898 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2899 if the connection is using TLS or if the remote IP is the same as the
2900 local IP, and the "valid-client-cert option" if a client certificate has
2903 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2904 authenticators. This can be used for authorization after authentication
2905 succeeds. (In the case of plaintext, it servers for both authentication
2908 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2909 if any retry times were supplied.
2911 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2912 connection, where there is no client IP address to check. In this
2913 situation, the verify now always succeeds.
2915 PH/25 Applied John Jetmore's -Mset patch.
2917 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2919 PH/27 In a string expansion for a processed (not raw) header when multiple
2920 headers of the same name were present, leading whitespace was being
2921 removed from all of them, but trailing whitespace was being removed only
2922 from the last one. Now trailing whitespace is removed from each header
2923 before concatenation. Completely empty headers in a concatenation (as
2924 before) are ignored.
2926 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2927 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2929 PH/29 [Removed. This was a change that I later backed out, and forgot to
2930 correct the ChangeLog entry (that I had efficiently created) before
2931 committing the later change.]
2933 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2934 address errors (4xx response to RCPT) over the same connection as other
2935 messages routed to the same hosts. Such deliveries are always "forced",
2936 so retry times are not inspected. This resulted in far too many retries
2937 for the affected addresses. The effect occurred only when there were more
2938 hosts than the hosts_max_try setting in the smtp transport when it had
2939 the 4xx errors. Those hosts that it had tried were not added to the list
2940 of hosts for which the message was waiting, so if all were tried, there
2941 was no problem. Two fixes have been applied:
2943 (i) If there are any address or message errors in an SMTP delivery, none
2944 of the hosts (tried or untried) are now added to the list of hosts
2945 for which the message is waiting, so the message should not be a
2946 candidate for sending over the same connection that was used for a
2947 successful delivery of some other message. This seems entirely
2948 reasonable: after all the message is NOT "waiting for some host".
2949 This is so "obvious" that I'm not sure why it wasn't done
2950 previously. Hope I haven't missed anything, but it can't do any
2951 harm, as the worst effect is to miss an optimization.
2953 (ii) If, despite (i), such a delivery is accidentally attempted, the
2954 routing retry time is respected, so at least it doesn't keep
2955 hammering the server.
2957 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2958 in ${readsocket because some servers need this prod.
2960 PH/32 Added some extra debug output when updating a wait-xxx database.
2962 PH/33 The hint "could be header name not terminated by colon", which has been
2963 given for certain expansion errors for a long time, was not being given
2964 for the ${if def:h_colon_omitted{... case.
2966 PH/34 The spec says: "With one important exception, whenever a domain list is
2967 being scanned, $domain contains the subject domain." There was at least
2968 one case where this was not true.
2970 PH/35 The error "getsockname() failed: connection reset by peer" was being
2971 written to the panic log as well as the main log, but it isn't really
2972 panic-worthy as it just means the connection died rather early on. I have
2973 removed the panic log writing for the ECONNRESET error when getsockname()
2976 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2977 runs only) independently of the message's sender address. This meant
2978 that, if the 4xx error was in fact related to the sender, a different
2979 message to the same recipient with a different sender could confuse
2980 things. In particular, this can happen when sending to a greylisting
2981 server, but other circumstances could also provoke similar problems.
2982 I have changed the default so that the retry time for these errors is now
2983 based a combination of the sender and recipient addresses. This change
2984 can be overridden by setting address_retry_include_sender=false in the
2987 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2988 remote server are returned as part of bounce messages. This was not
2989 happening for LMTP over a pipe (the lmtp transport), but now it is the
2990 same for both kinds of LMTP.
2992 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2993 in header lines that were in fact CNAMEs. This is no longer the case.
2995 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2996 and queue runs started by the daemon processed all messages. This has
2997 been fixed so that -R and -S can now usefully be given with -q<time>.
2999 PH/40 Import PCRE release 6.7 (fixes some bugs).
3001 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3003 PH/42 Give an error if -q is specified more than once.
3005 PH/43 Renamed the variables $interface_address and $interface_port as
3006 $received_ip_address and $received_port, to make it clear that these
3007 values apply to message reception, and not to the outgoing interface when
3008 a message is delivered. (The old names remain recognized, of course.)
3010 PH/44 There was no timeout on the connect() call when using a Unix domain
3011 socket in the ${readsocket expansion. There now is.
3013 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3014 be meaningful with "accept".
3016 SC/01 Eximstats V1.43
3017 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3019 SC/02 Eximstats V1.44
3020 Use a glob alias rather than an array ref in the generated
3021 parser. This improves both readability and performance.
3023 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3024 Collect SpamAssassin and rejection statistics.
3025 Don't display local sender or destination tables unless
3026 there is data to show.
3027 Added average volumes into the top table text output.
3029 SC/04 Eximstats V1.46
3030 Collect data on the number of addresses (recipients)
3031 as well as the number of messages.
3033 SC/05 Eximstats V1.47
3034 Added 'Message too big' to the list of mail rejection
3035 reasons (thanks to Marco Gaiarin).
3037 SC/06 Eximstats V1.48
3038 Mainlog lines which have GMT offsets and are too short to
3039 have a flag are now skipped.
3041 SC/07 Eximstats V1.49 (Alain Williams)
3042 Added the -emptyok flag.
3044 SC/08 Eximstats V1.50
3045 Fixes for obtaining the IP address from reject messages.
3047 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3048 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3049 whitespace changes from 4.64-PH/27
3051 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3054 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3055 are found, allow negative numbers in numeric criteria)
3057 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3059 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3060 to match changes made in 4.64-PH/43
3062 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3064 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3065 "domain" is actually an IP address, and force a failure. This locks out
3066 those revolvers/nameservers that support "A-for-A" lookups, in
3067 contravention of the specifications.
3069 PH/48 When a host name was looked up from an IP address, and the subsequent
3070 forward lookup of the name timed out, the host name was left in
3071 $sender_host_name, contrary to the specification.
3073 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3074 restricted to single-key lookups, Exim was not diagnosing an error if
3075 * or *@ was used with a query-style lookup.
3077 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3079 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3080 long ago, but noone interested enough thought of it. Let's just say that
3081 the "1.1" means that there are some new functions that weren't there at
3082 some point in the past.
3084 PH/51 Error processing for expansion failure of helo_data from an smtp
3085 transport during callout processing was broken.
3087 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3088 tested/used via the -bh/-bhc/-bs options.
3090 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3091 bug, fixed in subsequent PCRE releases).
3093 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3094 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3096 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3102 SC/01 Use a glob alias rather than an array ref in eximstats generated
3103 parser. This improves both readability and performance.
3105 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3106 Don't display local sender or destination tables in eximstats unless
3107 there is data to show.
3108 Added average volumes into the eximstats top table text output.
3110 SC/03 Collect data on the number of addresses (recipients) as well
3111 as the number of messages in eximstats.
3113 TF/01 Correct an error in the documentation for the redirect router. Exim
3114 does (usually) call initgroups() when daemonizing.
3116 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3117 with consistent privilege compared to when running as a daemon.
3119 TF/03 Note in the spec that $authenticated_id is not set for local
3120 submissions from trusted users.
3122 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3123 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3125 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3126 by adding some example configuration directives to the default
3127 configuration file. A little bit of work is required to uncomment the
3128 directives and define how usernames and passwords are checked, but
3129 there is now a framework to start from.
3131 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3132 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3133 without this. I don't know how relevant this is to other LDAP libraries.
3135 PH/02 Add the verb name to the "unknown ACL verb" error.
3137 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3139 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3141 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3142 directory not expanded when it should be if an expanded home directory
3143 was set for the address (which is overridden by the transport).
3145 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3148 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3149 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3150 because it is too late at that time, and has no effect.
3152 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3153 security issue with \' (bugzilla #107). I could not use the
3154 PQescapeStringConn() function, because it needs a PGconn value as one of
3157 PH/08 When testing addresses using -bt, indicate those final addresses that
3158 are duplicates that would not cause an additional delivery. At least one
3159 person was confused, thinking that -bt output corresponded to deliveries.
3160 (Suppressing duplicates isn't a good idea as you lose the information
3161 about possibly different redirections that led to the duplicates.)
3163 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3164 systems where poll() doesn't work, in particular OS X.
3166 PH/10 Added more information to debugging output for retry time not reached.
3168 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3169 operations in malware.c.
3171 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3174 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3175 syslog_duplication set false, log lines that would normally be written
3176 both the the main log and to the reject log were not written to syslog at
3179 PH/14 In the default configuration, change the use of "message" in ACL warn
3180 statements to "add_header".
3182 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3183 not followed by a command (e.g. "seen endif").
3185 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3186 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3189 PH/17 Added extra conditions to the default value of delay_warning_condition
3193 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3194 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3195 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3198 The Auto-Submitted: and various List- headers are standardised, whereas I
3199 don't think Precedence: ever was.
3201 PH/18 Refactored debugging code in route_finduser() to show more information,
3202 in particular, the error code if getpwnam() issues one.
3204 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3205 This is apparently needed in addition to the PH/07 change above to avoid
3206 any possible encoding problems.
3208 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3209 but not after initializing Perl.
3211 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3212 output them only if debugging. By default they are written stderr,
3213 apparently, which is not desirable.
3215 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3218 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3221 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3223 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3224 authenticated or an ident call has been made. Suppress the default
3225 values for $authenticated_id and $authenticated_sender (but permit -oMai
3226 and -oMas) when testing with -bh.
3228 PH/24 Re-jigged the order of the tests in the default configuration so that the
3229 tests for valid domains and recipients precede the DNS black list and CSA
3230 tests, on the grounds that those ones are more expensive.
3232 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3233 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3234 This bug exists in every version of Exim that I still have, right back to
3237 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3238 However, an attempt to turn on case-sensitivity in a regex key by
3239 including (?-i) didn't work because the subject string was already
3240 lowercased, and the effects were non-intuitive. It turns out that a
3241 one-line patch can be used to allow (?-i) to work as expected.
3247 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3248 other effects) broke the use of negated acl sub-conditions.
3250 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3253 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3254 "Deny" causes Exim to reject the incoming connection with a 554 error.
3255 Unfortunately, if there is a major crisis, such as a disk failure,
3256 tcp-wrappers gives "deny", whereas what one would like would be some
3257 kind of temporary error. A kludge has been added to help with this.
3258 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3259 554 error is used if errno is still zero or contains ENOENT (which occurs
3260 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3263 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3265 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3266 errors. Otherwise a message that provokes a temporary error (when other
3267 messages do not) can cause a whole host to time out.
3269 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3270 addresses were routed directly to files or pipes from a redirect router.
3271 File deliveries just didn't batch; pipe deliveries might have suffered
3274 PH/06 A failure to get a lock for a hints database would erroneously always say
3275 "Failed to get write lock", even when it was really a read lock.
3277 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3278 of 0600. This has been changed to use the value of the lockfile_mode
3279 option (which defaults to 0600).
3281 PH/08 Applied small patch from the Sieve maintainer.
3283 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3284 folder from quota calculations, a direct delivery into this folder messed
3285 up the contents of the maildirsize file. This was because the regex was
3286 used only to exclude .Trash (or whatever) when the size of the mailbox
3287 was calculated. There was no check that a delivery was happening into an
3288 excluded directory. This bug has been fixed by ignoring all quota
3289 processing for deliveries into excluded directories.
3291 PH/10 Added the maildirfolder_create_regex option to appendfile.
3297 PH/01 The code for finding all the local interface addresses on a FreeBSD
3298 system running IPv6 was broken. This may well have applied to all BSD
3299 systems, as well as to others that have similar system calls. The broken
3300 code found IPv4 interfaces correctly, but gave incorrect values for the
3301 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3302 that it would not match correctly against @[] and not recognize the IPv6
3305 PH/02 The ipliteral router was not recognizing addresses of the form user@
3306 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3308 PH/03 Added disable_ipv6.
3310 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3311 decoded form, because it is most often used to construct To: headers
3312 lines in autoreplies, and the decoded form may well be syntactically
3313 invalid. However, $reply_address has leading white space removed, and all
3314 newlines turned into spaces so that the autoreply transport does not
3317 PH/05 If group was specified without a user on a router, and no group or user
3318 was specified on a transport, the group from the router was ignored.
3320 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3321 for visible compile-time settings that can be used to change these
3322 numbers, for those that want even more. Backwards compatibility with old
3323 spool files has been maintained. However, going back to a previous Exim
3324 release will lost any variables that are in spool files.
3326 PH/07 Two small changes when running in the test harness: increase delay when
3327 passing a TCP/IP connection to a new process, in case the original
3328 process has to generate a bounce, and remove special handling of
3329 127.0.0.2 (sic), which is no longer necessary.
3331 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3332 be the same on different OS.
3334 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3337 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3338 whether --show-vars was specified or not
3340 JJ/02 exipick: Added support for new ACL variable spool format introduced
3343 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3344 syntactically invalid From: or Reply-to: line, and a filter used this to
3345 generate an autoreply, and therefore failed to obtain an address for the
3346 autoreply, Exim could try to deliver to a non-existent relative file
3347 name, causing unrelated and misleading errors. What now happens is that
3348 it logs this as a hard delivery error, but does not attempt to create a
3351 PH/11 The exinext utility has a -C option for testing purposes, but although
3352 the given file was scanned by exinext itself; it wasn't being passed on
3353 when Exim was called.
3355 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3356 an end-of-file indication when reading a command response.
3358 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3359 compiled. In many other places in Exim, IPv6 addresses are always
3360 recognized, so I have changed this. It also means that IPv4 domain
3361 literals of the form [IPV4:n.n.n.n] are now always recognized.
3363 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3364 used if the router is not running as root, for example, when verifying at
3365 ACL time, or when using -bh. The debugging output from this situation was
3366 non-existent - all you got was a failure to exec. I have made two
3369 (a) Failures to set uid/gid, the current directory, or a process leader
3370 in a subprocess such as that created by queryprogram now generate
3371 suitable debugging output when -d is set.
3373 (b) The queryprogram router detects when it is not running as root,
3374 outputs suitable debugging information if -d is set, and then runs
3375 the subprocess without attempting to change uid/gid.
3377 PH/15 Minor change to Makefile for building test_host (undocumented testing
3380 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3381 additional section of a DNS packet that returns MX or SRV records.
3382 Instead, it always explicitly searches for A/AAAA records. This avoids
3383 major problems that occur when a DNS server includes only records of one
3384 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3385 fixed another bug: if SRV records were looked up and the corresponding
3386 address records were *not* found in the additional section, the port
3387 values from the SRV records were lost.
3389 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3390 using the correct key (the original address) when searching the retry
3391 rules in order to find which one to use for generating the retry hint.
3393 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3394 adding the default one. Similarly, if it contains a Reply-To: header, the
3395 errors_reply_to option, if set, is not used.
3397 PH/19 When calculating a retry time, Exim used to measure the "time since
3398 failure" by looking at the "first failed" field in the retry record. Now
3399 it does not use this if it is later than than the arrival time of the
3400 message. Instead it uses the arrival time. This makes for better
3401 behaviour in cases where some deliveries succeed, thus re-setting the
3402 "first failed" field. An example is a quota failure for a huge message
3403 when small messages continue to be delivered. Without this change, the
3404 "time since failure" will always be short, possible causing more frequent
3405 delivery attempts for the huge message than are intended.
3406 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3408 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3409 $1, $2, $3) because the numerical variables can be reset during some
3410 expansion items (e.g. "match"), thereby losing the authentication data.
3412 PH/21 Make -bV show the size of off_t variables so that the test suite can
3413 decide whether to run tests for quotas > 2G.
3415 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3416 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3417 filecount value is greater than 2G or if a quota value is greater than 2G
3418 on a system where the size of off_t is not greater than 4, a panic error
3421 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3422 never match. The debug and -bh output now contains an explicit error
3423 message indicating a malformed IPv4 address or mask.
3425 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3426 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3427 PH/23 above applies.
3429 PH/25 Do not write to syslog when running in the test harness. The only
3430 occasion when this arises is a failure to open the main or panic logs
3431 (for which there is an explicit test).
3433 PH/26 Added the /no_tell option to "control=freeze".
3435 PH/27 If a host name lookup failed very early in a connection, for example, if
3436 the IP address matched host_lookup and the reverse lookup yielded a name
3437 that did not have a forward lookup, an error message of the form "no IP
3438 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3439 could be logged. Now it outputs the IP address instead of "NULL".
3441 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3442 allows the sender and the authenticated sender to be set when
3443 submitting a message from within Exim. Since child_open_exim() is
3444 documented for local_scan(), the new function should be too.
3446 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3447 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3448 results in an empty string is now treated as unset.
3450 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3452 PH/31 Added sender_verify_fail logging option.
3454 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3455 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3456 tidied the source and removed it altogether.
3458 PH/33 When a queue run was abandoned because the load average was too high, a
3459 log line was always written; now it is written only if the queue_run log
3460 selector is set. In addition, the log line for abandonment now contains
3461 information about the queue run such as the pid. This is always present
3462 in "start" and "stop" lines but was omitted from the "abandon" line.
3464 PH/34 Omit spaces between a header name and the colon in the error message that
3465 is given when verify = headers_syntax fails (if there are lots of them,
3466 the message gets confusing).
3468 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3469 names, as there are now some PTR records that contain slashes. This check
3470 is only to protect against broken name servers that fall over on strange
3471 characters, so the fact that it applies to all lookups doesn't matter.
3473 PH/36 Now that the new test suite is complete, we can remove some of the
3474 special code in Exim that was needed for the old test suite. For example,
3475 sorting DNS records because real resolvers return them in an arbitrary
3476 order. The new test suite's fake resolver always returns records in the
3479 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3480 bounces) except when queue_only is set, to avoid logging races between
3481 the different processes.
3483 PH/38 Panic-die if .include specifies a non-absolute path.
3485 PH/39 A tweak to the "H" retry rule from its user.
3487 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3488 a label. They prevented compilation on older perls.
3490 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3491 a warning to be raised on newish perls.
3493 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3494 on queue. Changes to match documented behaviour of showing count of
3495 messages matching specified criteria.
3497 PH/40 Changed the default ident timeout from 30s to 5s.
3499 PH/41 Added support for the use of login_cap features, on those BSD systems
3500 that have them, for controlling the resources used by pipe deliveries.
3502 PH/42 The content-scanning code uses fopen() to create files in which to put
3503 message data. Previously it was not paying any attention to the mode of
3504 the files. Exim runs with umask(0) because the rest of the code creates
3505 files with open(), and sets the required mode explicitly. Thus, these
3506 files were ending up world-writeable. This was not a big issue, because,
3507 being within the spool directory, they were not world-accessible. I have
3508 created a function called modefopen, which takes an additional mode
3509 argument. It sets umask(777), creates the file, chmods it to the required
3510 mode, then resets the umask. All the relevant calls to fopen() in the
3511 content scanning code have been changed to use this function.
3513 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3514 to 24 hours. This avoids potential overflow problems when processing G
3515 and H retry rules. I suspect nobody ever tinkers with this value.
3517 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3519 PH/45 When the plaintext authenticator is running as a client, the server's
3520 challenges are checked to ensure they are valid base64 strings. By
3521 default, the authentication attempt is cancelled if an invalid string is
3522 received. Setting client_ignore_invalid_base64 true ignores these errors.
3523 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3524 they are received. Thus, the responses can be made to depend on the
3525 challenges. If an invalid string is ignored, an empty string is placed in
3528 PH/46 Messages that are created by the autoreply transport now contains a
3529 References: header, in accordance with RFCs 2822 and 3834.
3531 PH/47 Added authenticated_sender_force to the smtp transport.
3533 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3535 PH/49 Installed latest patch from the Sieve maintainer.
3537 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3538 was also set, the appendfile transport was unnecessarily scanning a
3539 directory of message files (e.g. for maildir delivery) to find the count
3540 of files (along with the size), even though it did not need this
3541 information. It now does the scan only if it needs to find either the
3542 size of the count of files.
3544 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3546 PH/52 Two bugs concerned with error handling when the smtp transport is
3549 (i) Exim was not creating retry information for temporary errors given
3550 for individual recipients after the DATA command when the smtp transport
3551 was used in LMTP mode. This meant that they could be retried too
3552 frequently, and not timed out correctly.
3554 (ii) Exim was setting the flag that allows error details to be returned
3555 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3556 recipients that were returned after the DATA command.
3558 PH/53 This is related to PH/52, but is more general: for any failing address,
3559 when detailed error information was permitted to be returned to the
3560 sender, but the error was temporary, then after the final timeout, only
3561 "retry timeout exceeded" was returned. Now it returns the full error as
3562 well as "retry timeout exceeded".
3564 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3565 do this, and (what is worse) MTAs that accept it.
3567 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3568 will now be deprecated.
3570 PH/56 New os.c-cygwin from the Cygwin maintainer.
3572 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3573 formats (previously only available in exim formats via -bpr, -bpru,
3574 and -bpra. Now also available in native and exiqgrep formats)
3576 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3577 with very large, slow to parse queues
3579 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3581 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3583 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3584 responses to authentication challenges, though it was showing the
3585 challenges; (ii) I've removed the CR characters from the debug output for
3588 PH/58 Allow for the insertion of a newline as well as a space when a string
3589 is turned into more than one encoded-word during RFC 2047 encoding. The
3590 Sieve code now uses this.
3592 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3593 data_4xx, lost_connection, tls_required.
3595 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3596 message was being sent as an SMTP response.
3598 PH/61 Add -l and -k options to exicyclog.
3600 PH/62 When verifying, if an address was redirected to one new address, so that
3601 verification continued, and the new address failed or deferred after
3602 having set something in $address_data, the value of $address_data was not
3603 passed back to the ACL. This was different to the case when no
3604 redirection occurred. The value is now passed back in both cases.
3606 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3607 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3608 use login_cap.h, so on its own it isn't the distinguishing feature. The
3609 new name refers directly to the setclassresources() function.
3611 PH/65 Added configuration files for NetBSD3.
3613 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3615 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3616 is preferred over IPv4.
3618 PH/68 The bounce_return_message and bounce_return_body options were not being
3619 honoured for bounces generated during the reception of non-SMTP messages.
3620 In particular, this applied to messages rejected by the ACL. This bug has
3621 been fixed. However, if bounce_return_message is true and bounce_return_
3622 body is false, the headers that are returned for a non-SMTP message
3623 include only those that have been read before the error was detected.
3624 (In the case of an ACL rejection, they have all been read.)
3626 PH/69 The HTML version of the specification is now built in a directory called
3627 spec_html instead of spec.html, because the latter looks like a path with
3628 a MIME-type, and this confuses some software.
3630 PH/70 Catch two compiler warnings in sieve.c.
3632 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3633 function verify_get_ident() calls ip_connect() to connect a socket, but
3634 if the "connect()" function timed out, ip_connect() used to close the
3635 socket. However, verify_get_ident() also closes the socket later, and in
3636 between Exim writes to the log, which may get opened at this point. When
3637 the socket was closed in ip_connect(), the log could get the same file
3638 descriptor number as the socket. This naturally causes chaos. The fix is
3639 not to close the socket in ip_connect(); the socket should be closed by
3640 the function that creates it. There was only one place in the code where
3641 this was missing, in the iplookup router, which I don't think anybody now
3642 uses, but I've fixed it anyway.
3644 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3645 well as to direct DNS lookups. Otherwise the handling of names in host
3646 lists is inconsistent and therefore confusing.
3652 PH/01 Two changes to the default runtime configuration:
3654 (1) Move the checks for relay_from_hosts and authenticated clients from
3655 after to before the (commented out) DNS black list checks.
3657 (2) Add control=submission to the relay_from_hosts and authenticated
3658 clients checks, on the grounds that messages accepted by these
3659 statements are most likely to be submissions.
3661 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3663 (1) Generate an error if the third argument for the ${prvs expansion is
3666 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3669 (3) Reset the variables that are obtained from the first argument of
3670 ${prvscheck and used in the second argument before leaving the code,
3671 because their memory is reclaimed, so using them afterwards may do
3674 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3675 one (it's much easier than Tom thought :-).
3677 (5) Because of (4), we can now allow for the use of $prvscheck_result
3678 inside the third argument.
3680 PH/03 For some reason, the default setting of PATH when running a command from
3681 a pipe transport was just "/usr/bin". I have changed it to
3684 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3685 anything to be listed in the output from -bV.
3687 PH/05 When a filter generated an autoreply, the entire To: header line was
3688 quoted in the delivery log line, like this:
3690 => >A.N.Other <ano@some.domain> <original@ddress> ...
3692 This has been changed so that it extracts the operative address. There
3693 may be more than one such address. If so, they are comma-separated, like
3696 => >ano@some.domain,ona@other.domain <original@ddress> ...
3698 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3699 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3700 not being looked up in the rDNS to get a host name, Exim was showing the
3701 IP address twice in Received: lines, even though the IP addresses were
3702 identical. For example:
3704 Received: from [1.2.3.4] (helo=[1.2.3.4])
3706 However, if the real host name was known, it was omitting the HELO data
3707 if it matched the actual IP address. This has been tidied up so that it
3708 doesn't show the same IP address twice.
3710 PH/07 When both +timestamp and +memory debugging was on, the value given by
3711 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3712 called by the debug printing, thereby overwriting the timestamp buffer.
3713 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3715 PH/08 When the original message was included in an autoreply transport, it
3716 always said "this is a copy of the message, including all the headers",
3717 even if body_only or headers_only was set. It now gives an appropriate
3720 PH/09 Applied a patch from the Sieve maintainer which:
3722 o fixes some comments
3723 o adds the (disabled) notify extension core
3724 o adds some debug output for the result of if/elsif tests
3725 o points to the current vacation draft in the documentation
3726 and documents the missing references header update
3730 o fixes a bug in processing the envelope test (when testing
3731 multiple envelope elements, the last element determined the
3734 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3735 Electronic Mail") by including:
3737 Auto-submitted: auto-generated
3739 in the messages that it generates (bounce messages and others, such as
3740 warnings). In the case of bounce messages for non-SMTP messages, there was
3741 also a typo: it was using "Auto_submitted" (underscore instead of
3742 hyphen). Since every message generated by Exim is necessarily in response
3743 to another message, thes have all been changed to:
3745 Auto-Submitted: auto-replied
3747 in accordance with these statements in the RFC:
3749 The auto-replied keyword:
3751 - SHOULD be used on messages sent in direct response to another
3752 message by an automatic process,
3754 - MUST NOT be used on manually-generated messages,
3756 - MAY be used on Delivery Status Notifications (DSNs) and Message
3757 Disposition Notifications (MDNs),
3759 - MUST NOT be used on messages generated by automatic or periodic
3760 processes, except for messages which are automatic responses to
3763 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3764 to the default Received: header definition.
3766 PH/12 Added log selector acl_warn_skipped (default on).
3768 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3769 variables because (a) they are in the wrong storage pool and (b) even if
3770 they were copied, it wouldn't work properly because of the caching.
3772 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3773 checking when decoding. Apparently there are clients that generate
3774 overlong encoded strings. Why am I not surprised?
3776 PH/15 If the first argument of "${if match_address" was not empty, but did not
3777 contain an "@" character, Exim crashed. Now it writes a panic log message
3778 and treats the condition as false.
3780 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3782 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3783 extension "envelope-auth". The code is finished and in agreement with
3784 other implementations, but there is no documentation so far and in fact,
3785 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3786 not changing the active code.
3788 Print executed "if" and "elsif" statements when debugging is used. This
3789 helps a great deal to understand what a filter does.
3791 Document more things not specified clearly in RFC3028. I had all this
3792 sorted out, when out of a sudden new issues came to my mind. Oops."
3794 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3797 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3798 canonical form (as documented). However, after a host name lookup from
3799 the IP address, check_host() was doing a simple string comparison with
3800 addresses acquired from the DNS when checking that the found name did
3801 have the original IP as one of its addresses. Since any found IPv6
3802 addresses are likely to be in abbreviated form, the comparison could
3803 fail. Luckily, there already exists a function for doing the comparison
3804 by converting both addresses to binary, so now that is used instead of
3805 the text comparison.
3807 PH/20 There was another similar case to PH/19, when a complete host name was
3808 given in a host list; looking up its IP address could give an abbreviated
3809 form, whereas the current host's name might or might not be abbreviated.
3810 The same fix has been applied.
3816 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3817 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3820 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3821 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3823 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3825 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3826 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3827 Also, for "Auto-Submitted", treat anything other than "no" as
3828 non-personal, in accordance with RFC 3834. (Previously it treated
3829 anything starting "auto-" as non-personal.)
3831 TF/01 The control=submission/name=... option had a problem with syntax
3832 errors if the name included a slash character. The /name= option
3833 now slurps the rest of the string, so it can include any characters
3834 but it must come last in the list of options (after /sender_retain
3837 PH/05 Some modifications to the interface to the fake nameserver for the new
3845 TK/01 Added the "success_on_redirect" address verification option. See
3846 NewStuff for rationale and an example.
3848 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3850 PH/02 Patch to exigrep to allow it to work on syslog lines.
3852 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3853 fread() to skip over the body file's header line, because in Cygwin the
3854 header line is locked and is inaccessible.
3856 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3857 co-exist for some time) to make it clear that it is the Exim ID that is
3858 referenced, not the Message-ID: header line.
3860 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3861 string_format() function, because snprintf() does not exist on all
3864 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3865 ${stat: expansion item.
3867 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3868 protocol synchronization error", to keep the pedants happy.
3870 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3871 well as for IRIX systems, when gcc is being used. See the host.c source
3874 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3876 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3879 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3880 transport and to the smtp transport in LMTP mode.
3882 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3884 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3885 run a filter in a subprocess. This could lead to confusion in subsequent
3886 lookups in the parent process. There should also be a search_tidyup() at
3887 the end of the subprocess.
3889 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3890 only if the host matched helo_try_verify_hosts, which caused the
3891 verification to occur when the EHLO/HELO command was issued. The ACL just
3892 tested the remembered result. Now, if a previous verification attempt has
3893 not happened, "verify = helo" does it there and then.
3895 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3897 TK/03 Fix log output including CR from clamd.
3899 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3900 exist provoked a memory error which could cause a segfault.
3902 PH/15 Installed PCRE 6.2
3904 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3906 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3907 of the problem. Specifically, suggested +O2 rather than +O1 for the
3910 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3912 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3913 record, the port was not correctly logged when the outgoing_port log
3914 selector was set (it logged the transort's default port).
3916 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3917 fallback_hosts, and "hosts" in the smtp transport.
3919 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3920 host errors such as "Connection refused".
3922 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3923 authentication with radiusclient 0.4.9:
3925 - Error returned from rc_read_config was caught wrongly
3926 - Username/password not passed on to radius server due to wrong length.
3928 The presumption is that some radiusclient API changes for 4.51/PH/17
3929 were not taken care of correctly. The code is still untested by me (my
3930 Linux distribution still has 0.3.2 of radiusclient), but it was
3931 contributed by a Radius user.
3933 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3934 expanding the "port" option of the smtp transport.
3936 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3937 while reading a MIME header. Thanks to Tom Hughes for a patch.
3939 PH/24 Include config.h inside local_scan.h so that configuration settings are
3942 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3943 that in an ACL for RCPT (for example), you can examine exactly what was
3946 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3947 commands, but it was not correctly comparing the address with the actual
3948 client host address. Thus, it would show the EHLO address in Received:
3949 header lines when this was not necessary.
3951 PH/27 Added the % operator to ${eval:}.
3953 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3954 it should be ignoring failures (because with -C, for example, it has lost
3955 privilege). It wasn't ignoring creation failures other than "already
3958 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3961 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3962 because an input line was too long, either on its own, or by virtue of
3963 too many continuations, the temporary file was not being removed, and the
3964 return code was incorrect.
3966 PH/31 Missing "BOOL" in function definition in filtertest.c.
3968 PH/32 Applied Sieve patches from the maintainer.
3970 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3972 PH/33 Added "verify = not_blind".
3974 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3975 Local/Makefile (with some defaults set). These are used in built scripts
3976 such as exicyclog, but they have never been used in the exim_install
3977 script (though there are many overriding facilities there). I have
3978 arranged that the exim_install script now takes note of these two
3981 PH/35 Installed configuration files for Dragonfly.
3983 PH/36 When a locally submitted message by a trusted user did not contain a
3984 From: header, and the sender address was obtained from -f or from an SMTP
3985 MAIL command, and the trusted user did not use -F to supply a sender
3986 name, $originator_name was incorrectly used when constructing a From:
3987 header. Furthermore, $originator_name was used for submission mode
3988 messages from external hosts without From: headers in a similar way,
3989 which is clearly wrong.
3991 PH/37 Added control=suppress_local_fixups.
3993 PH/38 When log_selector = +received_sender was set, and the addition of the
3994 sender made the log line's construction buffer exactly full, or one byte
3995 less than full, an overflow happened when the terminating "\n" was
3998 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3999 when the result of a list match is failure because a DNS lookup failed.
4001 PH/40 RM_COMMAND is now used in the building process.
4003 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4004 the "build-* directories that it finds.
4006 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4007 address in a domain literal was a prefix of an interface address.
4009 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4010 when verifying a sender address, unless rewrite_headers is false.
4012 PH/44 Wrote a long comment about why errors_to addresses are verified as
4013 recipients, not senders.
4015 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4016 the ratelimit ACL was added.
4018 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4020 PH/46 Added extra information about PostgreSQL errors to the error string.
4022 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4023 avoiding the need to install special zones in a real server. This is
4024 backwards compatible; if it can't find the fake resolver, it drops back.
4025 Thus, both old and new test suites can be run.
4027 TF/02 Added util/ratelimit.pl
4029 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4030 clock is set back in time.
4032 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4033 Candler <B.Candler@pobox.com>.
4035 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4036 OK for addresses that are the result of redirections.
4038 PH/48 A number of further additions for the benefit of the new test suite,
4039 including a fake gethostbyname() that interfaces to the fake DNS resolver
4042 TF/06 The fix for widen_domains has also been applied to qualify_single and
4043 search_parents which are the other dnslookup options that can cause
4046 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4049 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4051 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4052 service" if it encountered an MX record with an empty target hostname.
4053 The message is now "an MX or SRV record indicated no SMTP service".
4055 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4056 if the DNS of the sending site is misconfigured. This is quite a
4057 common situation. This change restores the behaviour of treating a
4058 helo verification defer as a failure.
4060 PH/51 If self=fail was set on a router, the bounce message did not include the
4061 actual error message.
4067 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4069 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4070 command itself ended in error, the underlying message about the transport
4071 filter timeout was being overwritten with the pipe command error. Now the
4072 underlying error message should be appended to the second error message.
4074 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4076 PH/02 Reduce the amount of output that "make" produces by default. Full output
4077 can still be requested.
4079 PH/03 The warning log line about a condition test deferring for a "warn" verb
4080 was being output only once per connection, rather than after each
4081 occurrence (because it was using the same function as for successful
4082 "warn" verbs). This seems wrong, so I have changed it.
4084 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4085 it should not have, which might have caused a crash in the right
4086 circumstances, but probably never did.
4088 PH/04 Installed a modified version of Tony Finch's patch to make submission
4089 mode fix the return path as well as the Sender: header line, and to
4090 add a /name= option so that you can make the user's friendly name appear
4093 TF/03 Added the control = fakedefer ACL modifier.
4095 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4096 Mark Lowes for thorough testing.
4098 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4100 TK/03 Merged latest SRS patch from Miles Wilton.
4102 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4103 with the definition in sysexits.h (which is #included earlier).
4104 Fortunately, Exim does not actually use EX_OK. The code used to try to
4105 preserve the sysexits.h value, by assuming that macro definitions were
4106 scanned for macro replacements. I have been disabused of this notion,
4107 so now the code just undefines EX_OK before #including unistd.h.
4109 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4110 in the smtp transport. When a block could not be written in a single
4111 write() function, the timeout was being re-applied to each part-write.
4112 This seems wrong - if the receiver was accepting one byte at a time it
4113 would take for ever. The timeout is now adjusted when this happens. It
4114 doesn't have to be particularly precise.
4116 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4117 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4119 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4120 without a domain if the check to <postmaster@domain> fails.
4122 SC/01 Eximstats: added -xls and the ability to specify output files
4123 (patch written by Frank Heydlauf).
4125 SC/02 Eximstats: use FileHandles for outputting results.
4127 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4129 SC/04 Eximstats: fixed display of large numbers with -nvr option
4131 SC/05 Eximstats: fixed merging of reports with empty tables.
4133 SC/06 Eximstats: added the -include_original_destination flag
4135 SC/07 Eximstats: removed tabs and trailing whitespace.
4137 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4139 TK/06 MBOX spool code: Add real "From " MBOX separator line
4140 so the .eml file is really in mbox format (even though
4141 most programs do not really care). Patch from Alex Miller.
4143 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4144 The latter is generated from $received_to and is only set if the
4145 message has one envelope recipient. SA can use these headers,
4146 obviously out-of-the-box. Patch from Alex Miller.
4148 PH/08 The ${def test on a variable was returning false if the variable's
4149 value was "0", contrary to what the specification has always said!
4150 The result should be true unless the variable is empty.
4152 PH/09 The syntax error of a character other than { following "${if
4153 def:variable_name" (after optional whitespace) was not being diagnosed.
4154 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4155 accidental colon was present, for example, could give incorrect results.
4157 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4158 result is used (not including appendfile, where other changes are about
4161 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4162 This involved changing a lot of size variables from int to off_t. It
4163 should work with maildirs and everything.
4165 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4166 spamd dying while we are connected to it.
4168 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4171 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4172 function for BDB 4.3.
4174 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4176 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4177 the sources for PCRE have been re-arranged and more files are now
4180 PH/15 The code I had for printing potentially long long variables in PH/11
4181 above was not the best (it lost precision). The length of off_t variables
4182 is now inspected at build time, and an appropriate printing format (%ld
4183 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4184 to be "long long int" or "long int". This is needed for the internal
4185 formatting function string_vformat().
4187 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4188 the configuration file to be ":syslog", then the script "guesses" where
4189 the logs files are, rather than using the compiled in default. In our
4190 case the guess is not the same as the compiled default, so the script
4191 suddenly stopped working when I started to use syslog. The patch checks
4192 to see if log_file_path is "". If so, it attempts to read it from exim
4193 with no configuration file to get the compiled in version, before it
4194 falls back to the previous guessing code."
4196 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4197 implementing BATV in an Exim configuration. See NewStuff for the gory
4200 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4201 Makefile that are specific to HP-UX.
4203 PH/18 If the "use_postmaster" option was set for a recipient callout together
4204 with the "random" option, the postmaster address was used as the MAIL
4205 FROM address for the random test, but not for the subsequent recipient
4206 test. It is now used for both.
4208 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4209 patch removes a few documentation additions to RFC 3028, because the
4210 latest draft now contains them. It adds the new en;ascii-case comparator
4211 and a new error check for 8bit text in MIME parts. Comparator and
4212 require names are now matched exactly. I enabled the subaddress
4213 extension, but it is not well tested yet (read: it works for me)."
4215 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4216 rework some of the code of TK/09 above to avoid the hardwired use of
4217 "%lld" and "long long". Replaced the call to snprintf() with a call to
4220 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4221 records point to non-existent hosts", "retry timeout exceeded", and
4222 "retry time not reached for any host after a long failure period".
4224 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4225 experimental DomainKeys support:
4227 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4228 (2) On an error such as an illegally used "control", the wrong name for
4229 the control was given.
4231 These problems did NOT occur unless DomainKeys support was compiled.
4233 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4235 PH/24 Added ${if match_ip condition.
4237 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4238 databases so that it will be absolutely obvious if a crash occurs in the
4239 DB library. This is a regular occurrence (often caused by mis-matched
4242 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4243 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4244 picked up on a user's system that detects such things. There doesn't seem
4245 to be a gcc warning option for this - only an attribute that has to be
4246 put on the function's prototype. It seems that in Fedora Core 4 they have
4247 set this on a number of new functions. No doubt there will be more in due
4250 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4251 specify a transport. However, if an address that was verified by such a
4252 router was the subject of a callout, Exim crashed because it tried to
4253 read the rcpt_include_affixes from the non-existent transport. Now it
4254 just assumes that the setting of that option is false. This bug was
4255 introduced by 4.51/PH/31.
4257 PH/28 Changed -d+all to exclude +memory, because that information is very
4258 rarely of interest, but it makes the output a lot bigger. People tend to
4259 do -d+all out of habit.
4261 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4262 code in os-type was giving problems when libc.so lives in lib64, like on
4265 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4266 aren't the modern standard, and it seems that some systems' include files
4267 don't always have them. Exim was already checking for some of the newer
4268 ones like T_AAAA, and defining it itself. I've added checks for all the
4269 record types that Exim uses.
4271 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4272 not automatically generating a new one, as it is supposed to. This
4273 prevented TLS from working. If the file did exist, but contained invalid
4274 data, a new version was generated, as expected. It was only the case of a
4275 non-existent file that was broken.
4277 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4278 with a change in libdomainkeys > 0.64.
4280 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4281 from DNS. If the selector record carries the flag, it now has
4282 precedence over the domain-wide flag.
4284 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4286 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4287 the use of a port name that isn't defined in /etc/services) occurred, the
4288 message was deferred as in a normal delivery, and thus remained on the
4289 spool, instead of being failed because of the mua_wrapper setting. This
4290 is now fixed, and I tidied up some of the mua_wrapper messages at the
4293 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4294 the messages in a hash of arrays rather than using individual hashes.
4295 This is a bit cleaner and results in dramatic memory savings, albeit
4296 at a slight CPU cost.
4298 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4299 as requested by Marc Sherman.
4301 SC/10 Eximstats: added histograms for user specified patterns as requested
4304 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4306 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4307 fopen() in the content-scanning modules that did not already have it.
4313 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4314 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4316 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4318 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4320 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4321 to test Sieve filters that use "vacation".
4323 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4324 that changes the way the GnuTLS parameters are stored in the cache file.
4325 The new format can be generated externally. For backward compatibility,
4326 if the data in the cache doesn't make sense, Exim assumes it has read an
4327 old-format file, and it generates new data and writes a new file. This
4328 means that you can't go back to an older release without removing the
4331 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4332 work if there are any delivery delays because "one_time" forces the
4333 parent to be marked "delivered", so its unseen clone is never tried
4334 again. For this reason, Exim now forbids the simultaneous setting of
4337 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4338 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4339 entry if you want to know the details. The fix, however, neglected to
4340 consider the case when local delivery batching is involved. The test for
4341 "previously delivered" was not happening when checking to see if an
4342 address could be batched with a previous (undelivered) one; under
4343 certain circumstances this could lead to multiple deliveries to the same
4346 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4347 in its include files, and this causes problems building Exim.
4349 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4350 header_syntax) but Exim was just ignoring anything given after a slash.
4351 In particular, this caused confusion with an attempt to use "verify =
4352 reverse_host_lookup/defer_ok". An error is now given when options are
4353 supplied for verify items that do not have them. (Maybe reverse_host_
4354 lookup should have a defer_ok option, but that's a different point.)
4356 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4357 defined by RFC 821) to 2048, because there were problems with some AUTH
4358 commands, and RFC 1869 says the size should be increased for extended
4359 SMTP commands that take arguments.
4361 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4364 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4365 "unknown" error; now it says that the functionality isn't in the binary.
4367 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4368 an address' error message when a string expansion fails (syntax or
4369 whatever). Otherwise the password may appear in the log. Following change
4370 PH/42 below, there is no longer a chance of it appearing in a bounce
4373 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4375 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4376 of its addresses was ever tried. (Bugzilla bug #2.)
4378 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4379 the result incorrectly in the debug output. (It correctly added a newline
4380 to what was transported.)
4382 TF/01 Added $received_time.
4384 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4385 commented out examples of how to interface to a virus scanner and to
4386 SpamAssassin. Also added commented examples of av_scanner and
4387 spamd_address settings.
4389 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4390 and controls are allowed in which ACLs. There were a couple of minor
4391 errors. Some of the entries in the conditions table (which is a table of
4392 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4393 negation of where the condition IS allowed.
4395 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4397 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4398 header file does not have a version number, so I've had to invent a new
4399 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4400 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4401 radiusclient), but it was contributed by a Radius user.
4403 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4404 files or directories, for OpenSSL.
4406 PH/19 When an Exim process that is running as root has to create an Exim log
4407 file, it does so in a subprocess that runs as exim:exim so as to get the
4408 ownership right at creation (otherwise, other Exim processes might see
4409 the file with the wrong ownership). There was no test for failure of this
4410 fork() call, which would lead to the process getting stuck as it waited
4411 for a non-existent subprocess. Forks do occasionally fail when resources
4412 run out. I reviewed all the other calls to fork(); they all seem to check
4415 PH/20 When checking for unexpected SMTP input at connect time (before writing
4416 the banner), Exim was not dealing correctly with a non-positive return
4417 from the read() function. If the client had disconnected by this time,
4418 the result was a log entry for a synchronization error with an empty
4419 string after "input=" when read() returned zero. If read() returned -1
4420 (an event I could not check), uninitialized data bytes were printed.
4421 There were reports of junk text (parts of files, etc) appearing after
4424 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4426 PH/22 Added support for macro redefinition, and (re)definition in between
4427 driver and ACL definitions.
4429 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4430 forgetting to use the resulting value; it was using the unexpanded value.
4432 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4433 hadn't been configured. The fix is from Juergen Kreileder, who
4434 understands it better than I do:
4436 "Here's what I see happening with three configured cyrus_sasl
4437 authenticators configured (plain, login, cram-md5):
4439 On startup auth_cyrus_sasl_init() gets called for each of these.
4440 This means three calls to sasl_listmech() without a specified mech_list.
4441 => SASL tests which mechs of all available mechs actually work
4442 => three warnings about OTP not working
4443 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4445 With the patch, sasl_listmech() also gets called three times. But now
4446 SASL's mech_list option is set to the server_mech specified in the the
4447 authenticator. Or in other words, the answer from sasl_listmech()
4448 gets limited to just the mech you're testing for (which is different
4450 => the return list contains just 'plain' or 'login', 'cram-md5' or
4451 nothing depending on the value of ob->server_mech.
4453 I've just tested the patch: Authentication still works fine,
4454 unavailable mechs specified in the exim configuration are still
4455 caught, and the auth.log warnings about OTP are gone."
4457 PH/25 When debugging is enabled, the contents of the command line are added
4458 to the debugging output, even when log_selector=+arguments is not
4461 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4462 answer is "GNU", and only if the return is "GNU/something" is the answer
4465 PH/27 $acl_verify_message is now set immediately after the failure of a
4466 verification in an ACL, and so is available in subsequent modifiers. In
4467 particular, the message can be preserved by coding like this:
4469 warn !verify = sender
4470 set acl_m0 = $acl_verify_message
4472 Previously, $acl_verify_message was set only while expanding "message"
4473 and "log_message" when a very denied access.
4475 PH/28 Modified OS/os.c-Linux with
4477 -#ifndef OS_LOAD_AVERAGE
4478 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4480 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4481 nomenclature these days.)
4483 PH/29 Installed patch from the Sieve maintainer that adds the options
4484 sieve_useraddress and sieve_subaddress to the redirect router.
4486 PH/30 In these circumstances:
4487 . Two addresses routed to the same list of hosts;
4488 . First host does not offer TLS;
4489 . First host accepts first address;
4490 . First host gives temporary error to second address;
4491 . Second host offers TLS and a TLS session is established;
4492 . Second host accepts second address.
4493 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4494 and peerdn, if requested) that were in fact used only for the second
4497 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4498 attention to any local part prefix or suffix that was matched by the
4499 router that accepted the address. It now behaves in the same way as it
4500 does for delivery: the affixes are removed from the local part unless
4501 rcpt_include_affixes is set on the transport.
4503 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4504 timeout during the DATA phase of an incoming message.
4506 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4507 applied a patch sanctioned by the Sieve maintainer.
4509 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4510 the uid or gid is negative. A case of a negative gid caused this to be
4511 noticed. The fix allows for either to be negative.
4513 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4514 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4517 PH/36 The change PH/12 above was broken. Fixed it.
4519 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4520 the grounds that routing the same address twice would always produce the
4521 same answer. This might have been true once, but it is certainly no
4522 longer true now. Routing a child address may depend on the previous
4523 routing that produced that child. Some complicated redirection strategies
4524 went wrong when messages had multiple recipients, and made Exim's
4525 behaviour dependent on the order in which the addresses were given.
4527 I have moved the duplicate checking until after the routing is complete.
4528 Exim scans the addresses that are assigned to local and remote
4529 transports, and removes any duplicates. This means that more work will be
4530 done, as duplicates will always all be routed, but duplicates are
4531 presumably rare, so I don't expect this is of any significance.
4533 For deliveries to pipes, files, and autoreplies, the duplicate checking
4534 still happens during the routing process, since they are not going to be
4537 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4538 It corrects a timeout issue with spamd. This is Ian's comment: "The
4539 background is that sometimes spamd either never reads data from a
4540 connection it has accepted, or it never writes response data. The exiscan
4541 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4542 blindly assumes that writes won't block so it may never time out."
4544 PH/39 Allow G after quota size as well as K and M.
4546 PH/40 The value set for $authenticated_id in an authenticator may not contain
4547 binary zeroes or newlines because the value is written to log lines and
4548 to spool files. There was no check on this. Now the value is run through
4549 the string_printing() function so that such characters are converted to
4550 printable escape sequences.
4552 PH/41 $message_linecount is a new variable that contains the total number of
4553 lines in the message. Compare $body_linecount, which is the count for the
4556 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4557 bounce and delay warning messages, except in certain special cases, which
4560 (a) An SMTP error message from a remote host;
4561 (b) A message specified in a :fail: redirection;
4562 (c) A message specified in a "fail" command in a system filter;
4563 (d) A message specified in a FAIL return from the queryprogram router;
4564 (e) A message specified by the cannot_route_message router option.
4566 In these cases only, Exim does include the error details in bounce and
4567 warning messages. There are also a few cases where bland messages such
4568 as "unrouteable address" or "local delivery error" are given.
4570 PH/43 $value is now also set for the "else" part of a ${run expansion.
4572 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4573 being worked on, but at least Exim now implements the latest version to
4576 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4577 process to complete was treated as a delivery failure, a timeout while
4578 writing the message to the pipe was logged, but erroneously treated as a
4579 successful delivery. Such timeouts include transport filter timeouts. For
4580 consistency with the overall process timeout, these timeouts are now
4581 treated as errors, giving rise to delivery failures by default. However,
4582 there is now a new Boolean option for the pipe transport called
4583 timeout_defer, which, if set TRUE, converts the failures into defers for
4584 both kinds of timeout. A transport filter timeout is now identified in
4587 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4588 systems where "make" and "gmake" are different, calling "gmake" at top
4589 level broke things. I've arranged for the value of $(MAKE) to be passed
4590 from the Makefile to this script so that it can call the same version of
4594 A note about Exim versions 4.44 and 4.50
4595 ----------------------------------------
4597 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4598 changes of various kinds. As a consequence, a big documentation update was
4599 needed. This delayed the release for rather longer than seemed good, especially
4600 in the light of a couple of (minor) security issues. Therefore, the changes
4601 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4602 release. So 4.44 and 4.50 are in effect two different branches that both start
4605 I have left the 4.50 change log unchanged; it contains all the changes since
4606 4.43. The change log for 4.44 is below; many of its items are identical to
4607 those for 4.50. This seems to be the most sensible way to preserve the
4608 historical information.
4614 1. Minor wording change to the doc/README.SIEVE file.
4616 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4617 computation of the current number of files was incorrect.
4619 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4620 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4623 4. Give more explanation in the error message when the command for a transport
4624 filter fails to execute.
4626 5. There are several places where Exim runs a non-Exim command in a
4627 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4628 was being done only for the command run by the queryprogram router. It is
4629 now done for all such subprocesses. The other cases are: ${run, transport
4630 filters, and the commands run by the lmtp and pipe transports.
4632 6. Added CONFIGURE_GROUP build-time option.
4634 7. Some older OS have a limit of 256 on the maximum number of file
4635 descriptors. Exim was using setrlimit() to set 1000 as a large value
4636 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4637 systems. I've change it so that if it can't get 1000, it tries for 256.
4639 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4640 was an oversight, and furthermore, ever since the addition of extra
4641 controls (e.g. 4.43/32), the checks on when to allow different forms of
4642 "control" were broken. There should now be diagnostics for all cases when a
4643 control that does not make sense is encountered.
4645 9. Added the /retain_sender option to "control=submission".
4647 10. $recipients is now available in the predata ACL (oversight).
4649 11. Tidy the search cache before the fork to do a delivery from a message
4650 received from the command line. Otherwise the child will trigger a lookup
4651 failure and thereby defer the delivery if it tries to use (for example) a
4652 cached ldap connection that the parent has called unbind on.
4654 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4655 of $address_data from the recipient verification was clobbered by the
4656 sender verification.
4658 13. The value of address_data from a sender verification is now available in
4659 $sender_address_data in subsequent conditions in the ACL statement.
4661 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4663 15. Added a new option "connect=<time>" to callout options, to set a different
4666 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4667 was its contents. (It was OK if the option was not defined at all.)
4669 17. A "Completed" log line is now written for messages that are removed from
4670 the spool by the -Mrm option.
4672 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4673 information about exactly what failed.
4675 19. Added -dd to debug only the daemon process.
4677 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4678 handles timeouts, both on the server side and network timeouts. Renamed the
4679 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4681 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4682 It is now set to "smtps".
4684 22. $host_address is now set to the target address during the checking of
4685 ignore_target_hosts.
4687 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4688 being passed; this would have caused $sender_host_name to have been used if
4689 matching the list had actually called for a host name (not very likely,
4690 since this list is usually IP addresses). A host name is now passed as
4693 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4694 code that specifies a non-restarting handler (typically sigaction() in
4695 modern systems) in an attempt to fix a rare and obscure crash bug.
4697 25. Narrowed the window for a race in the daemon that could cause it to ignore
4698 SIGCHLD signals. This is not a major problem, because they are used only to
4699 wake it up if nothing else does.
4701 26. A malformed maildirsize file could cause Exim to calculate negative values
4702 for the mailbox size or file count. Odd effects could occur as a result.
4703 The maildirsize information is now recalculated if the size or filecount
4706 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4707 support for a long time. Removed HAVE_SYS_VFS_H.
4709 28. Installed the latest version of exipick from John Jetmore.
4711 29. In an address list, if the pattern was not a regular expression, an empty
4712 subject address (from a bounce message) matched only if the pattern was an
4713 empty string. Non-empty patterns were not even tested. This was the wrong
4714 because it is perfectly reasonable to use an empty address as part of a
4715 database query. An empty address is now tested by patterns that are
4716 lookups. However, all the other forms of pattern expect the subject to
4717 contain a local part and a domain, and therefore, for them, an empty
4718 address still always fails if the pattern is not itself empty.
4720 30. Exim went into a mad DNS loop when attempting to do a callout where the
4721 host was specified on an smtp transport, and looking it up yielded more
4722 than one IP address.
4724 31. Re-factored the code for checking spool and log partition space into a
4725 function that finds that data and another that does the check. The former
4726 is then used to implement four new variables: $spool_space, $log_space,
4727 $spool_inodes, and $log_inodes.
4729 32. The RFC2047 encoding function was originally intended for short strings
4730 such as real names; it was not keeping to the 75-character limit for
4731 encoded words that the RFC imposes. It now respects the limit, and
4732 generates multiple encoded words if necessary. To be on the safe side, I
4733 have increased the buffer size for the ${rfc2047: expansion operator from
4736 33. It is now permitted to omit both strings after an "if" condition; if the
4737 condition is true, the result is "true". As before, when the second string
4738 is omitted, a false condition yields an empty string. This makes it less
4739 cumbersome to write custom ACL and router conditions.
4741 34. Failure to deliver a bounce message always caused it to be frozen, even if
4742 there was an errors_to setting on the router. The errors_to setting is now
4745 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4746 canonical form (fully expanded) before being placed in
4747 $sender_host_address.
4749 36. The table in the code that translates DNS record types into text (T_A to
4750 "A" for instance) was missing entries for NS and CNAME. It is just possible
4751 that this could have caused confusion if both these types were looked up
4752 for the same domain, because the text type is used as part of Exim's
4753 per-process caching. But the chance of anyone hitting this buglet seems
4756 37. The dnsdb lookup has been extended in a number of ways.
4758 (1) There is a new type, "zns", which walks up the domain tree until it
4759 finds some nameserver records. It should be used with care.
4761 (2) There is a new type, "mxh", which is like "mx" except that it returns
4762 just the host names, not the priorities.
4764 (3) It is now possible to give a list of domains (or IP addresses) to be
4765 looked up. The behaviour when one of the lookups defers can be
4766 controlled by a keyword.
4768 (4) It is now possible to specify the separator character for use when
4769 multiple records are returned.
4771 38. The dnslists ACL condition has been extended: it is now possible to supply
4772 a list of IP addresses and/or domains to be looked up in a particular DNS
4775 39. Added log_selector=+queue_time_overall.
4777 40. When running the queue in the test harness, wait just a tad after forking a
4778 delivery process, to get repeatability of debugging output.
4780 41. Include certificate and key file names in error message when GnuTLS fails
4781 to set them up, because the GnuTLS error message doesn't include the name
4782 of the failing file when there is a problem reading it.
4784 42. Allow both -bf and -bF in the same test run.
4786 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4788 44. The "Exiscan patch" is now merged into the mainline Exim source.
4790 45. Sometimes the final signoff response after QUIT could fail to get
4791 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4792 < 0 before doing a fflush(). This bug looks as though it goes back to the
4793 introduction of TLS in release 3.20, but "sometimes" must have been rare
4794 because the tests only now provoked it.
4796 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4797 (this can affect the format of dates).
4799 47. exim_tidydb, when checking for the continued existence of a message for
4800 which it has found a message-specific retry record, was not finding
4801 messages that were in split spool directories. Consequently, it was
4802 deleting retry records that should have stayed in existence.
4804 48. Steve fixed some bugs in eximstats.
4806 49. The SPA authentication driver was not abandoning authentication and moving
4807 on to the next authenticator when an expansion was forced to fail,
4808 contradicting the general specification for all authenticators. Instead it
4809 was generating a temporary error. It now behaves as specified.
4811 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4812 (the order specifies the preference for clients). The order is now AES256,
4813 AES128, 3DES, ARCFOUR128.
4815 51. Small patch to Sieve code - explicitly set From: when generating an
4818 52. Exim crashed if a remote delivery caused a very long error message to be
4819 recorded - for instance if somebody sent an entire SpamAssassin report back
4820 as a large number of 550 error lines. This bug was coincidentally fixed by
4821 increasing the size of one of Exim's internal buffers (big_buffer) that
4822 happened as part of the Exiscan merge. However, to be on the safe side, I
4823 have made the code more robust (and fixed the comments that describe what
4826 53. Now that there can be additional text after "Completed" in log lines (if
4827 the queue_time_overall log selector is set), a one-byte patch to exigrep
4828 was needed to allow it to recognize "Completed" as not the last thing in
4831 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4832 patch that reportedly fixes this has been added. I am not expert enough to
4833 create a test for it. This is what the patch creator wrote:
4835 "I found a little strange behaviour of ldap code when working with
4836 Windows 2003 AD Domain, where users was placed in more than one
4837 Organization Units. When I tried to give exim partial DN, the exit code
4838 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4839 But simultaneously result of request was absolutely normal ldap result,
4840 so I produce this patch..."
4842 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4843 REFERENCE, so I have modified the code to exclude the patch when that macro
4846 55. Some experimental protocols are using DNS PTR records for new purposes. The
4847 keys for these records are domain names, not reversed IP addresses. The
4848 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4849 leaves it alone. Component reversal etc. now happens only for IP addresses.
4852 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4854 57. Double the size of the debug message buffer (to 2048) so that more of very
4855 long debug lines gets shown.
4857 58. The exicyclog utility now does better if the number of log files to keep
4858 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4860 59. Two changes related to the smtp_active_hostname option:
4862 (1) $smtp_active_hostname is now available as a variable.
4863 (2) The default for smtp_banner uses $smtp_active_hostname instead
4864 of $primary_hostname.
4866 60. The host_aton() function is supposed to be passed a string that is known
4867 to be a valid IP address. However, in the case of IPv6 addresses, it was
4868 not checking this. This is a hostage to fortune. Exim now panics and dies
4869 if the condition is not met. A case was found where this could be provoked
4870 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4871 components; fortuitously, this particular loophole had already been fixed
4872 by change 4.50/55 above.
4874 If there are any other similar loopholes, the new check in host_aton()
4875 itself should stop them being exploited. The report I received stated that
4876 data on the command line could provoke the exploit when Exim was running as
4877 exim, but did not say which command line option was involved. All I could
4878 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4879 running as the user.
4882 61. There was a buffer overflow vulnerability in the SPA authentication code
4883 (which came originally from the Samba project). I have added a test to the
4884 spa_base64_to_bits() function which I hope fixes it.
4887 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4888 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4890 63. The daemon start-up calls getloadavg() while still root for those OS that
4891 need the first call to be done as root, but it missed one case: when
4892 deliver_queue_load_max is set with deliver_drop_privilege. This is
4893 necessary for the benefit of the queue runner, because there is no re-exec
4894 when deliver_drop_privilege is set.
4896 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4897 This has been fixed.
4899 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4900 was in use, was not putting the data itself into the right store pool;
4901 consequently, it could be overwritten for a subsequent message in the same
4902 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4905 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4907 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4908 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4909 the function were treating the return as a boolean value, which happened to
4910 work because 0=false and not-0=true, but is not correct code.
4912 68. The host_aton() function was not handling scoped IPv6 addresses (those
4913 with, for example, "%eth0" on the end) correctly.
4915 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4916 negated items (that is, ~something) in unsigned ints. Some compilers
4917 apparently mutter when there is no cast.
4919 70. If an address verification called from an ACL failed, and did not produce a
4920 user-specific message (i.e. there was only a "system" message), nothing was
4921 put in $acl_verify_message. In this situation, it now puts the system
4924 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4925 session; change 4.31/43 added the unwanted input to the log line - except
4926 that it did not do this in the start of session case. It now does.
4928 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4929 This is wrong and can cause the other end to generate a synchronization
4930 error if it is another Exim or anything else that does the synchronization
4931 check. A QUIT command is no longer sent after a timeout.
4933 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4934 during host lookups.
4936 74. The defer_ok option of callout verification was not working if it was used
4937 when verifying addresses in header lines, that is, for this case:
4939 verify = header_sender/callout=defer_ok
4941 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4942 those file descriptors could be used for SMTP connections. If anything
4943 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4944 could be sent to the SMTP client, causing chaos. The daemon now opens
4945 stdin, stdout, and stderr to /dev/null when it puts itself into the
4948 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4949 log by default. The user can override this with suitable Perl magic.
4951 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4952 the log message when discard triggers, was not working for the DATA ACL or
4953 for the non-SMTP ACL.
4955 78. Error message wording change in sieve.c.
4957 79. If smtp_accept_max_per_host was set, the number of connections could be
4958 restricted to fewer than expected, because the daemon was trying to set up
4959 a new connection before checking whether the processes handling previous
4960 connections had finished. The check for completed processes is now done
4961 earlier. On busy systems, this bug wouldn't be noticed because something
4962 else would have woken the daemon, and it would have reaped the completed
4965 80. If a message was submitted locally by a user whose login name contained one
4966 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4967 It caused a spool format error. I have fixed the spool reading code. A
4968 related problem was that the "from" clause in the Received: line became
4969 illegal because of the space(s). It is now covered by ${quote_local_part.
4971 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4974 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4976 83. Patch from Sieve maintainer to fix unterminated string problem in
4977 "vacation" handling.
4979 84. Some minor changes to the Linux configuration files to help with other
4980 OS variants using glibc.
4982 85. One more patch for Sieve to update vacation handling to latest spec.
4985 ----------------------------------------------------
4986 See the note above about the 4.44 and 4.50 releases.
4987 ----------------------------------------------------
4993 1. Change 4.43/35 introduced a bug that caused file counts to be
4994 incorrectly computed when quota_filecount was set in an appendfile
4997 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4998 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5001 3. Give more explanation in the error message when the command for a transport
5002 filter fails to execute.
5004 4. There are several places where Exim runs a non-Exim command in a
5005 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5006 was being done only for the command run by the queryprogram router. It is
5007 now done for all such subprocesses. The other cases are: ${run, transport
5008 filters, and the commands run by the lmtp and pipe transports.
5010 5. Some older OS have a limit of 256 on the maximum number of file
5011 descriptors. Exim was using setrlimit() to set 1000 as a large value
5012 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5013 systems. I've change it so that if it can't get 1000, it tries for 256.
5015 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5016 was an oversight, and furthermore, ever since the addition of extra
5017 controls (e.g. 4.43/32), the checks on when to allow different forms of
5018 "control" were broken. There should now be diagnostics for all cases when a
5019 control that does not make sense is encountered.
5021 7. $recipients is now available in the predata ACL (oversight).
5023 8. Tidy the search cache before the fork to do a delivery from a message
5024 received from the command line. Otherwise the child will trigger a lookup
5025 failure and thereby defer the delivery if it tries to use (for example) a
5026 cached ldap connection that the parent has called unbind on.
5028 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5029 of $address_data from the recipient verification was clobbered by the
5030 sender verification.
5032 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5033 was its contents. (It was OK if the option was not defined at all.)
5035 11. A "Completed" log line is now written for messages that are removed from
5036 the spool by the -Mrm option.
5038 12. $host_address is now set to the target address during the checking of
5039 ignore_target_hosts.
5041 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5042 being passed; this would have caused $sender_host_name to have been used if
5043 matching the list had actually called for a host name (not very likely,
5044 since this list is usually IP addresses). A host name is now passed as
5047 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5048 code that specifies a non-restarting handler (typically sigaction() in
5049 modern systems) in an attempt to fix a rare and obscure crash bug.
5051 15. Narrowed the window for a race in the daemon that could cause it to ignore
5052 SIGCHLD signals. This is not a major problem, because they are used only to
5053 wake it up if nothing else does.
5055 16. A malformed maildirsize file could cause Exim to calculate negative values
5056 for the mailbox size or file count. Odd effects could occur as a result.
5057 The maildirsize information is now recalculated if the size or filecount
5060 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5061 support for a long time. Removed HAVE_SYS_VFS_H.
5063 18. Updated exipick to current release from John Jetmore.
5065 19. Allow an empty sender to be matched against a lookup in an address list.
5066 Previously the only cases considered were a regular expression, or an
5069 20. Exim went into a mad DNS lookup loop when doing a callout where the
5070 host was specified on the transport, if the DNS lookup yielded more than
5073 21. The RFC2047 encoding function was originally intended for short strings
5074 such as real names; it was not keeping to the 75-character limit for
5075 encoded words that the RFC imposes. It now respects the limit, and
5076 generates multiple encoded words if necessary. To be on the safe side, I
5077 have increased the buffer size for the ${rfc2047: expansion operator from
5080 22. Failure to deliver a bounce message always caused it to be frozen, even if
5081 there was an errors_to setting on the router. The errors_to setting is now
5084 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5085 canonical form (fully expanded) before being placed in
5086 $sender_host_address.
5088 24. Updated eximstats to version 1.33
5090 25. Include certificate and key file names in error message when GnuTLS fails
5091 to set them up, because the GnuTLS error message doesn't include the name
5092 of the failing file when there is a problem reading it.
5094 26. Expand error message when OpenSSL has problems setting up cert/key files.
5097 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5098 (this can affect the format of dates).
5100 28. exim_tidydb, when checking for the continued existence of a message for
5101 which it has found a message-specific retry record, was not finding
5102 messages that were in split spool directories. Consequently, it was
5103 deleting retry records that should have stayed in existence.
5105 29. eximstats updated to version 1.35
5106 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5107 1.35 - bugfix such that pie charts by volume are generated correctly
5109 30. The SPA authentication driver was not abandoning authentication and moving
5110 on to the next authenticator when an expansion was forced to fail,
5111 contradicting the general specification for all authenticators. Instead it
5112 was generating a temporary error. It now behaves as specified.
5114 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5115 (the order specifies the preference for clients). The order is now AES256,
5116 AES128, 3DES, ARCFOUR128.
5118 31. Small patch to Sieve code - explicitly set From: when generating an
5121 32. Exim crashed if a remote delivery caused a very long error message to be
5122 recorded - for instance if somebody sent an entire SpamAssassin report back
5123 as a large number of 550 error lines. This bug was coincidentally fixed by
5124 increasing the size of one of Exim's internal buffers (big_buffer) that
5125 happened as part of the Exiscan merge. However, to be on the safe side, I
5126 have made the code more robust (and fixed the comments that describe what
5129 33. Some experimental protocols are using DNS PTR records for new purposes. The
5130 keys for these records are domain names, not reversed IP addresses. The
5131 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5132 leaves it alone. Component reversal etc. now happens only for IP addresses.
5135 34. The host_aton() function is supposed to be passed a string that is known
5136 to be a valid IP address. However, in the case of IPv6 addresses, it was
5137 not checking this. This is a hostage to fortune. Exim now panics and dies
5138 if the condition is not met. A case was found where this could be provoked
5139 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5140 components; fortuitously, this particular loophole had already been fixed
5141 by change 4.50/55 or 4.44/33 above.
5143 If there are any other similar loopholes, the new check in host_aton()
5144 itself should stop them being exploited. The report I received stated that
5145 data on the command line could provoke the exploit when Exim was running as
5146 exim, but did not say which command line option was involved. All I could
5147 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5148 running as the user.
5151 35. There was a buffer overflow vulnerability in the SPA authentication code
5152 (which came originally from the Samba project). I have added a test to the
5153 spa_base64_to_bits() function which I hope fixes it.
5156 36. The daemon start-up calls getloadavg() while still root for those OS that
5157 need the first call to be done as root, but it missed one case: when
5158 deliver_queue_load_max is set with deliver_drop_privilege. This is
5159 necessary for the benefit of the queue runner, because there is no re-exec
5160 when deliver_drop_privilege is set.
5162 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5163 was in use, was not putting the data itself into the right store pool;
5164 consequently, it could be overwritten for a subsequent message in the same
5165 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5168 38. Sometimes the final signoff response after QUIT could fail to get
5169 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5170 < 0 before doing a fflush(). This bug looks as though it goes back to the
5171 introduction of TLS in release 3.20, but "sometimes" must have been rare
5172 because the tests only now provoked it.
5178 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5179 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5180 Now it returns an integer. A number of calls were still expecting a T/F
5181 return. Fortuitously, in all cases, the tests worked in OK situations,
5182 which is the norm. However, things would have gone wrong on any write
5183 failures on the smtp file descriptor. This function is used when sending
5184 messages over SMTP and also when doing verify callouts.
5186 2. When Exim is called to do synchronous delivery of a locally submitted
5187 message (the -odf or -odi options), it no longer closes stderr before doing
5190 3. Implemented the mua_wrapper option.
5192 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5194 5. Implemented the functions header_remove(), header_testname(),
5195 header_add_at_position(), and receive_remove_recipient(), and exported them
5198 6. If an ACL "warn" statement specified the addition of headers, Exim already
5199 inserted X-ACL-Warn: at the start if there was no header name. However, it
5200 was not making this test for the second and subsequent header lines if
5201 there were newlines in the string. This meant that an invalid header could
5202 be inserted if Exim was badly configured.
5204 7. Allow an ACL "warn" statement to add header lines at the start or after all
5205 the Received: headers, as well as at the end.
5207 8. Added the rcpt_4xx retry error code.
5209 9. Added postmaster_mailfrom=xxx to callout verification option.
5211 10. Added mailfrom=xxxx to the callout verification option, for verify=
5214 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5215 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5217 12. Inserted some casts to stop certain compilers warning when using pointer
5218 differences as field lengths or precisions in printf-type calls (mostly
5219 affecting debugging statements).
5221 13. Added optional readline() support for -be (dynamically loaded).
5223 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5224 same clock tick as a message's arrival, so that its received time was the
5225 same as the "first fail" time on the retry record, and that message
5226 remained on the queue past the ultimate address timeout, every queue runner
5227 would try a delivery (because it was past the ultimate address timeout) but
5228 after another failure, the ultimate address timeout, which should have then
5229 bounced the address, did not kick in. This was a "< instead of <=" error;
5230 in most cases the first failure would have been in the next clock tick
5231 after the received time, and all would be well.
5233 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5234 being recognized when the domain list was tested by the match_domain
5235 condition in an expansion string.
5237 16. Added the ${str2b64: operator.
5239 17. Exim was always calling setrlimit() to set a large limit for the number of
5240 processes, without checking whether the existing limit was already
5241 adequate. (It did check for the limit on file descriptors.) Furthermore,
5242 errors from getrlimit() and setrlimit() were being ignored. Now they are
5243 logged to the main and panic logs, but Exim does carry on, to try to do its
5244 job under whatever limits there are.
5246 18. Imported PCRE 5.0.
5248 19. Trivial typo in log message " temporarily refused connection" (the leading
5251 20. If the log selector return_path_on_delivery was set and an address was
5252 redirected to /dev/null, the delivery process crashed because it assumed
5253 that a return path would always be set for a "successful" delivery. In this
5254 case, the whole delivery is bypassed as an optimization, and therefore no
5257 21. Internal re-arrangement: the function for sending a challenge and reading
5258 a response while authentication was assuming a zero-terminated challenge
5259 string. It's now changed to take a pointer and a length, to allow for
5260 binary data in such strings.
5262 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5264 23. Exim was not respecting finduser_retries when seeking the login of the
5265 uid under which it was called; it was always trying 10 times. (The default
5266 setting of finduser_retries is zero.) Also, it was sleeping after the final
5267 failure, which is pointless.
5269 24. Implemented tls_on_connect_ports.
5271 25. Implemented acl_smtp_predata.
5273 26. If the domain in control=submission is set empty, Exim assumes that the
5274 authenticated id is a complete email address when it generates From: or
5275 Sender: header lines.
5277 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5278 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5279 chown and chgrp in /bin and hostname in /usr/bin.
5281 28. Exim was keeping the "process log" file open after each use, just as it
5282 does for the main log. This opens the possibility of it remaining open for
5283 long periods when the USR1 signal hits a daemon. Occasional processlog
5284 errors were reported, that could have been caused by this. Anyway, it seems
5285 much more sensible not to leave this file open at all, so that is what now
5288 29. The long-running daemon process does not normally write to the log once it
5289 has entered its main loop, and it closes the log before doing so. This is
5290 so that log files can straightforwardly be renamed and moved. However,
5291 there are a couple of unusual error situations where the daemon does write
5292 log entries, and I had neglected to close the log afterwards.
5294 30. The text of an SMTP error response that was received during a remote
5295 delivery was being truncated at 512 bytes. This is too short for some of
5296 the long messages that one sometimes sees. I've increased the limit to
5299 31. It is now possible to make retry rules that apply only when a message has a
5300 specific sender, in particular, an empty sender.
5302 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5303 it possible to be selective about when SMTP synchronization is enforced.
5305 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5307 32. Implemented hosts_connection_nolog.
5309 33. Added an ACL for QUIT.
5311 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5314 35. Added mailbox_size and mailbox_filecount to appendfile.
5316 36. Added control = no_multiline_responses to ACLs.
5318 37. There was a bug in the logic of the code that waits for the clock to tick
5319 in the case where the clock went backwards by a substantial amount such
5320 that the microsecond fraction of "now" was more than the microsecond
5321 fraction of "then" (but the whole seconds number was less).
5323 38. Added support for the libradius Radius client library this is found on
5324 FreeBSD (previously only the radiusclient library was supported).
5330 1. When certain lookups returned multiple values in the form name=value, the
5331 quoting of the values was not always being done properly. Specifically:
5332 (a) If the value started with a double quote, but contained no whitespace,
5334 (b) If the value contained whitespace other than a space character (i.e.
5335 tabs or newlines or carriage returns) it was not quoted.
5336 This fix has been applied to the mysql and pgsql lookups by writing a
5337 separate quoting function and calling it from the lookup code. The fix
5338 should probably also be applied to nisplus, ibase and oracle lookups, but
5339 since I cannot test any of those, I have not disturbed their existing code.
5341 2. A hit in the callout cache for a specific address caused a log line with no
5342 reason for rejecting RCPT. Now it says "Previous (cached) callout
5343 verification failure".
5345 3. There was an off-by-one bug in the queryprogram router. An over-long
5346 return line was truncated at 256 instead of 255 characters, thereby
5347 overflowing its buffer with the terminating zero. As well as fixing this, I
5348 have increased the buffer size to 1024 (and made a note to document this).
5350 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5351 when Exim is waiting for an SMTP response from a remote server, Exim
5352 restarts its select() call on the socket, thereby resetting its timeout.
5353 This is not a problem when such interrupts are rare. Somebody set up a cron
5354 job to run exiwhat every 2 minutes, which is less than the normal select()
5355 timeout (5 or 10 minutes). This meant that the select() timeout never
5356 kicked in because it was always reset. I have fixed this by comparing the
5357 time when an interrupt arrives with the time at the start of the first call
5358 to select(). If more time than the timeout has elapsed, the interrupt is
5359 treated as a timeout.
5361 5. Some internal re-factoring in preparation for the addition of Sieve
5362 extensions (by MH). In particular, the "personal" test is moved to a
5363 separate function, and given an option for scanning Cc: and Bcc: (which is
5364 not set for Exim filters).
5366 6. When Exim created an email address using the login of the caller as the
5367 local part (e.g. when creating a From: or Sender: header line), it was not
5368 quoting the local part when it contained special characters such as @.
5370 7. Installed new OpenBSD configuration files.
5372 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5373 try to make them clearer.
5375 9. Callout options, other than the timeout value, were being ignored when
5376 verifying sender addresses in header lines. For example, when using
5378 verify = header_sender/callout=no_cache
5380 the cache was (incorrectly) being used.
5382 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5383 only the Cygwin environment.
5385 11. When return_path_on_delivery was set as a log selector, if different remote
5386 addresses in the same message used different return paths and parallel
5387 remote delivery occurred, the wrong values would sometimes be logged.
5388 (Whenever a remote delivery process finished, the return path value from
5389 the most recently started remote delivery process was logged.)
5391 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5392 lines when AUTH and/or TLS are in use. This is the "received protocol"
5393 field. Exim used to use "asmtp" for authenticated SMTP, without any
5394 indication (in the protocol name) for TLS use. Now it follows the RFC and
5395 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5396 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5397 names appear in log lines as well as in Received: header lines.
5399 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5400 extensions, and comparison tests, and to fix some bugs.
5402 14. Changes to the "personal" filter test:
5404 (1) The test was buggy in that it was just doing the equivalent of
5405 "contains" tests on header lines. For example, if a user's address was
5406 anne@some.where, the "personal" test would incorrectly be true for
5408 To: susanne@some.where
5410 This test is now done by extracting each address from the header in turn,
5411 and checking the entire address. Other tests that are part of "personal"
5412 are now done using regular expressions (for example, to check local parts
5413 of addresses in From: header lines).
5415 (2) The list of non-personal local parts in From: addresses has been
5416 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5417 taken from the Sieve specification recommendations.
5419 (3) If the message contains any header line starting with "List-" it is
5420 treated as non-personal.
5422 (4) The test for "circular" in the Subject: header line has been removed
5423 because it now seems ill-conceived.
5425 15. Minor typos in src/EDITME comments corrected.
5427 16. Installed latest exipick from John Jetmore.
5429 17. If headers_add on a router specified a text string that was too long for
5430 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5431 of string_sprintf() is now avoided.
5433 18. $message_body_size was not set (it was always zero) when running the DATA
5434 ACL and the local_scan() function.
5436 19. For the "mail" command in an Exim filter, no default was being set for
5437 the once_repeat time, causing a random time value to be used if "once" was
5438 specified. (If the value happened to be <= 0, no repeat happened.) The
5439 default is now 0s, meaning "never repeat". The "vacation" command was OK
5440 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5441 (I found it when inspecting the code).
5443 20. There is now an overall timeout for performing a callout verification. It
5444 defaults to 4 times the callout timeout, which applies to individual SMTP
5445 commands during the callout. The overall timeout applies when there is more
5446 than one host that can be tried. The timeout is checked before trying the
5447 next host. This prevents very long delays if there are a large number of
5448 hosts and all are timing out (e.g. when the network connections are timing
5449 out). The value of the overall timeout can be changed by specifying an
5450 additional sub-option for "callout", called "maxwait". For example:
5452 verify = sender/callout=5s,maxwait=20s
5454 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5455 to the end before writing, but this should make it even safer).
5457 22. Exim was forgetting that it had advertised PIPELINING for the second and
5458 subsequent messages on an SMTP connection. It was also not resetting its
5459 memory on STARTTLS and an internal HELO.
5461 23. When Exim logs an SMTP synchronization error within a session, it now
5462 records whether PIPELINING has been advertised or not.
5464 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5465 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5466 rather than long int.
5468 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5470 26. Added the never_mail option to autoreply.
5476 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5477 crash if the getsockname() call failed; this can happen if a connection is
5478 closed very soon after it is established. The problem was simply in the
5479 order in which certain operations were done, causing Exim to try to write
5480 to the SMTP stream before it had set up the file descriptor. The bug has
5481 been fixed by making things happen in the correct order.
5487 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5488 before the connection was closed, thus losing the rejection response.
5490 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5491 some early Solaris releases, but causes trouble in current releases where
5492 socklen_t is defined.
5494 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5497 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5500 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5501 that releasing the top of it at the end releases what was used for sub-
5502 expansions (unless the block got too big). However, discard this block if
5503 the first thing is a variable or header, so that we can use its block when
5504 it is dynamic (useful for very large $message_headers, for example).
5506 6. Lookups now cache *every* query, not just the most recent. A new, separate
5507 store pool is used for this. It can be recovered when all lookup caches are
5508 flushed. Lookups now release memory at the end of their result strings.
5509 This has involved some general refactoring of the lookup sources.
5511 7. Some code has been added to the store_xxx() functions to reduce the amount
5512 of flapping under certain conditions.
5514 8. log_incoming_interface used to affect only the <= reception log lines. Now
5515 it causes the local interface and port to be added to several more SMTP log
5516 lines, for example "SMTP connection from", and rejection lines.
5518 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5520 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5522 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5523 could be overwritten at the end of the current message (or the start of a
5524 new message if it was set in a HELO ACL). The value is now preserved for
5525 the duration of the SMTP connection.
5527 12. If a transport had a headers_rewrite setting, and a matching header line
5528 contained an unqualified address, that address was qualified, even if it
5529 did not match any rewriting rules. The underlying bug was that the values
5530 of the flags that permit the existence of unqualified sender and recipient
5531 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5532 non-local messages, and by -bnq for local messages) were not being
5533 preserved with the message after it was received.
5535 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5536 "next input=" as part of the text comprising the host identity instead of
5537 the correct text. The code was using the same buffer for two different
5538 strings. However, depending on which order the printing function evaluated
5539 its arguments, the bug did not always show up. Under Linux, for example, my
5540 test suite worked just fine.
5542 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5543 doesn't work with some older versions of Perl. It has been changed to "my",
5544 which in any case is probably the better facility to use.
5546 15. A really picky compiler found some instances of statements for creating
5547 error messages that either had too many or two few arguments for the format
5550 16. The size of the buffer for calls to the DNS resolver has been increased
5551 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5552 for addresses that have a lot of PTR records. This alleviates a problem; it
5553 does not fully solve it.
5555 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5556 buffer now truncates the list and logs the incident, which is the same
5557 action as happens when Exim is looking up a host name and its aliases.
5558 Previously in this situation something unpredictable would happen;
5559 sometimes it was "internal error: store_reset failed".
5561 18. If a server dropped the connection unexpectedly when an Exim client was
5562 using GnuTLS and trying to read a response, the client delivery process
5563 crashed while trying to generate an error log message.
5565 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5566 string, for example:
5568 warn message = H1: something\nH2: something
5570 the text was added as a single header line from Exim's point of view
5571 though it ended up OK in the delivered message. However, searching for the
5572 second and subsequent header lines using $h_h2: did not work. This has been
5573 fixed. Similarly, if a system filter added multiple headers in this way,
5574 the routers could not see them.
5576 20. Expanded the error message when iplsearch is called with an invalid key to
5577 suggest using net-iplsearch in a host list.
5579 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5580 ACLs are no longer actually imposed (and a message to that effect is
5583 22. If a "gecos" field in a passwd entry contained escaped characters, in
5584 particular, if it contained a \" sequence, Exim got it wrong when building
5585 a From: or a Sender: header from that name. A second bug also caused
5586 incorrect handling when an unquoted " was present following a character
5587 that needed quoting.
5589 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5590 was not being matched caselessly.
5592 24. Arranged for all hyphens in the exim.8 source to be escaped with
5595 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5596 properly. Recipient callout cache records were still being keyed to include
5597 the sender, even when use_sender was set false. This led to far more
5598 callouts that were necessary. The sender is no longer included in the key
5599 when use_sender is false.
5601 26. Added "control = submission" modifier to ACLs.
5603 27. Added the ${base62d: operator to decode base 62 numbers.
5605 28. dnsdb lookups can now access SRV records.
5607 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5608 the configuration file.
5610 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5611 (-v) mode. This makes the output for a verbose queue run more intelligible.
5613 31. Added a use_postmaster feature to recipient callouts.
5615 32. Added the $body_zerocount variable, containing the number of binary zero
5616 bytes in the message body.
5618 33. The time of last modification of the "new" subdirectory is now used as the
5619 "mailbox time last read" when there is a quota error for a maildir
5622 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5624 35. Added +ignore_unknown as a special item in host lists.
5626 36. Code for decoding IPv6 addresses in host lists is now included, even if
5627 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5628 address was recognized as an IP address, but was then not correctly decoded
5629 into binary, causing unexpected and incorrect effects when compared with
5636 1. Very minor rewording of debugging text in manualroute to say "list of
5637 hosts" instead of "hostlist".
5639 2. If verify=header_syntax was set, and a header line with an unqualified
5640 address (no domain) and a large number of spaces between the end of the
5641 name and the colon was received, the reception process suffered a buffer
5642 overflow, and (when I tested it) crashed. This was caused by some obsolete
5643 code that should have been removed. The fix is to remove it!
5645 3. When running in the test harness, delay a bit after writing a bounce
5646 message to get a bit more predictability in the log output.
5648 4. Added a call to search_tidyup() just before forking a reception process. In
5649 theory, someone could use a lookup in the expansion of smtp_accept_max_
5650 per_host which, without the tidyup, could leave open a database connection.
5652 5. Added the variables $recipient_data and $sender_data which get set from a
5653 lookup success in an ACL "recipients" or "senders" condition, or a router
5654 "senders" option, similar to $domain_data and $local_part_data.
5656 6. Moved the writing of debug_print from before to after the "senders" test
5659 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5660 problems for message scanning, either using a data ACL, or using
5661 local_scan() because the Received: header was not generated till after they
5662 were called (in order to set the time as the time of reception completion).
5663 I have revised the way this works. The header is now generated after the
5664 body is received, but before the ACL or local_scan() are called. After they
5665 are run, the timestamp in the header is updated.
5671 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5672 before starting a queue runner without re-exec. This happened only when
5673 deliver_drop_privilege was set or when the Exim user was set to root. The
5674 effect of the bug was that timeouts during subsequent deliveries caused
5675 crashes instead of being properly handled. The handler is now left at its
5676 default (and expected) setting.
5678 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5679 message, again when deliver_drop_privilege is set or Exim is run as root.
5680 The bug described in (1) was not present in this case, but the tidying up
5681 of the other signals was missing. I have made the two cases consistent.
5683 3. The ignore_target_hosts setting on a manualroute router was being ignored
5684 for hosts that were looked up using the /MX notation.
5686 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5689 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5690 operated on the sender address. After changing the $sender_address to <>
5691 for the sender address verify, Exim was re-instated it as the original
5692 (before rewriting) address, but remembering that it had rewritten it, so it
5693 wasn't rewriting it again. This bug also had the effect of breaking the
5694 sender address verification caching when the sender address was rewritten.
5696 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5697 This has been changed so that if the ip literal address matches
5698 ignore_target_hosts, the router declines.
5700 7. Added expansion conditions match_domain, match_address, and match_local_
5701 part (NOT match_host).
5703 8. The placeholder for the Received: header didn't have a length field set.
5705 9. Added code to Exim itself and to exim_lock to test for a specific race
5706 condition that could lead to file corruption when using MBX delivery. The
5707 issue is with the lockfile that is created in /tmp. If this file is removed
5708 after a process has opened it but before that process has acquired a lock,
5709 there is the potential for a second process to recreate the file and also
5710 acquire a lock. This could lead to two Exim processes writing to the file
5711 at the same time. The added code performs the same test as UW imapd; it
5712 checks after acquiring the lock that its file descriptor still refers to
5713 the same named file.
5715 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5716 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5719 11. Added the smtp_active_hostname option. If used, this will typically be made
5720 to depend on the incoming interface address. Because $interface_address is
5721 not set up until the daemon has forked a reception process, error responses
5722 that can happen earlier (such as "too many connections") no longer contain
5725 12. If an expansion in a condition on a "warn" statement fails because a lookup
5726 defers, the "warn" statement is abandoned, and the next ACL statement is
5727 processed. Previously this caused the whole ACL to be aborted.
5729 13. Added the iplsearch lookup type.
5731 14. Added ident_timeout as a log selector.
5733 15. Added tls_certificate_verified as a log selector.
5735 16. Added a global option tls_require_ciphers (compare the smtp transport
5736 option of the same name). This controls incoming TLS connections.
5738 17. I finally figured out how to make tls_require_ciphers do a similar thing
5739 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5740 before starting the TLS session.
5742 18. Tabs are now shown as \t in -bP output.
5744 19. If the log selector return_path_on_delivery was set, Exim crashed when
5745 bouncing a message because it had too many Received: header lines.
5747 20. If two routers both had headers_remove settings, and the first one included
5748 a superfluous trailing colon, the final name in the first list and the
5749 first name in the second list were incorrectly joined into one item (with a
5750 colon in the middle).
5756 1. Added -C and -D options to the exinext utility, mainly to make it easier
5757 to include in the automated testing, but these could be helpful when
5758 multiple configurations are in use.
5760 2. The exinext utility was not formatting the output nicely when there was
5761 an alternate port involved in the retry record key, nor when there was a
5762 message id as well (for retries that were specific to a specific message
5763 and a specific host). It was also confused by IPv6 addresses, because of
5764 the additional colons they contain. I have fixed the IPv4 problem, and
5765 patched it up to do a reasonable job for IPv6.
5767 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5768 delivery, the log line now contains "pipelined" if PIPELINING was used.
5770 4. An SMTP transport process used to panic and die if the bind() call to set
5771 an explicit outgoing interface failed. This has been changed; it is now
5772 treated in the same way as a connect() failure.
5774 5. A reference to $sender_host_name in the part of a conditional expansion
5775 that was being skipped was still causing a DNS lookup. This no longer
5778 6. The def: expansion condition was not recognizing references to header lines
5779 that used bh_ and bheader_.
5781 7. Added the _cache feature to named lists.
5783 8. The code for checking quota_filecount in the appendfile transport was
5784 allowing one more file than it should have been.
5786 9. For compatibility with Sendmail, the command line option
5794 and sets the incoming protocol and host name (for trusted callers). The
5795 host name and its colon can be omitted when only the protocol is to be set.
5796 Note the Exim already has two private options, -pd and -ps, that refer to
5797 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5798 "s", but I don't think that's a major issue.
5800 10. A number of refactoring changes to the code, none of which should affect
5803 (a) The number of logging options was getting close to filling up the
5804 32-bit word that was used as a bit map. I have split them into two classes:
5805 those that are passed in the argument to log_write(), and those that are
5806 only ever tested independently outside of that function. These are now in
5807 separate 32-bit words, so there is plenty of room for expansion again.
5808 There is no change in the user interface or the logging behaviour.
5810 (b) When building, for example, log lines, the code previously used a
5811 macro that called string_cat() twice, in order to add two strings. This is
5812 not really sufficiently general. Furthermore, there was one instance where
5813 it was actually wrong because one of the argument was used twice, and in
5814 one call a function was used. (As it happened, calling the function twice
5815 did not affect the overall behaviour.) The macro has been replaced by a
5816 function that can join an arbitrary number of extra strings onto a growing
5819 (c) The code for expansion conditions now uses a table and a binary chop
5820 instead of a serial search (which was left over from when there were very
5821 few conditions). Also, it now recognizes conditions like "pam" even when
5822 the relevant support is not compiled in: a suitably worded error message is
5823 given if an attempt is made to use such a condition.
5825 11. Added ${time_interval:xxxxx}.
5827 12. A bug was causing one of the ddress fields not to be passed back correctly
5828 from remote delivery subprocesses. The field in question was not being
5829 subsequently used, so this caused to problems in practice.
5831 13. Added new log selectors queue_time and deliver_time.
5833 14. Might have fixed a bug in maildirsizefile handling that threw up
5834 "unexpected character" debug warnings, and recalculated the data
5835 unnecessarily. In any case, I expanded the warning message to give more
5838 15. Added the message "Restricted characters in address" to the statements in
5839 the default ACL that block characters like @ and % in local parts.
5841 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5842 Three changes have been made:
5844 (a) There was a serious bug; a negative response to MAIL caused the whole
5845 recipient domain to be cached as invalid, thereby blocking all messages
5846 to all local parts at the same domain, from all senders. This bug has
5847 been fixed. The domain is no longer cached after a negative response to
5848 MAIL if the sender used is not empty.
5850 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5853 (c) A new callout option, "use_sender" has been added for people who want
5854 the modified behaviour.
5860 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5863 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5864 indeed breaks things for older releases.
5866 3. Added additional logging to the case where there is a problem reading data
5867 from a filter that is running in a subprocess using a pipe, in order to
5868 try to track down a specific problem.
5870 4. Testing facility fudge: when running in the test harness and attempting
5871 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5872 getting "No route to host". Convert this to a timeout.
5874 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5877 6. Some OS don't have socklen_t but use size_t instead. This affects the
5878 fifth argument of getsockopt() amongst other things. This is now
5879 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5880 can be set for individual OS. I have set it for SunOS5, OSF1, and
5881 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5882 some earlier ones do not.
5884 7. Change 4.30/15 was not doing the test caselessly.
5886 8. The standard form for an IPv6 address literal was being rejected by address
5887 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5888 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5889 this, as well as the form without the "IPv6" on the front (but only when
5890 address literals are enabled, of course).
5892 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5894 10. Exim crashed if a message with an empty sender address specified by -f
5895 encountered a router with an errors_to setting. This could be provoked only
5896 by a command such as
5900 where an empty string was supplied; "<>" did not hit this bug.
5902 11. Installed PCRE release 4.5.
5904 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5905 remained set. It is now erased.
5907 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5908 times from message ids (which are base 36 rather than the normal 62).
5910 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5911 were being counted as actual protocol errors, and logged if the log
5912 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5913 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5914 and DATA following a set of rejected RCPTs do not count as protocol errors.
5915 In other words, Exim assumes they were pipelined, though this may not
5916 actually be the case. Of course, in all cases the client gets an
5917 appropriate error code.
5919 15. If a lookup fails in an ACL condition, a message about the failure may
5920 be available; it is used if testing the ACL cannot continue, because most
5921 such messages specify what the cause of the deferral is. However, some
5922 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5923 that caused an old message to be retained and used if a later statement
5924 caused a defer, replacing the real cause of the deferral.
5926 16. If an IP address had so many PTR records that the DNS lookup buffer
5927 was not large enough to hold them, Exim could crash while trying to process
5928 the truncated data. It now detects and logs this case.
5930 17. Further to 4.21/58, another change has been made: if (and only if) the
5931 first line of a message (the first header line) ends with CRLF, a bare LF
5932 in a subsequent header line has a space inserted after it, so as not to
5933 terminate the header.
5935 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5936 unnecessarily, used atoi() instead of strtol(), and didn't check the
5937 termination when getting file sizes from file names by regex.
5939 19. Completely re-implemented the support for maildirsize files, in the light
5940 of a number of problems with the previous contributed implementation
5941 (4.30/29). In particular:
5943 . If the quota is zero, the maildirsize file is maintained, but no quota is
5946 . If the maildir directory does not exist, it is created before any attempt
5947 to write a maildirsize file.
5949 . The quota value in the file is just a cache; if the quota is changed in
5950 the transport, the new value overrides.
5952 . A regular expression is available for excluding directories from the
5955 20. The autoreply transport checks the characters in options that define the
5956 message's headers; it allows continued headers, but it was checking with
5957 isspace() after an embedded newline instead of explicitly looking for a
5960 21. If all the "regular" hosts to which an address was routed had passed their
5961 expiry times, and had not reached their retry times, the address was
5962 bounced, even if fallback hosts were defined. Now Exim should go on to try
5965 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5966 equivalent code in the SMTP transport. Some hosts send humungous responses
5967 to HELO/EHLO, more than 1024 it seems.
5969 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5970 gives compiler warnings in some environments. I've now done it "properly",
5973 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5974 (because of problems with the built-in one) was declared to return uschar *
5975 instead of char *, causing compiler failure.
5977 25. Fixed a file descriptor leak when processing alias/forward files.
5979 26. Fixed a minor format string issue in dbfn.c.
5981 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5983 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5984 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5987 29. When a sender address is verified, it is cached, to save repeating the test
5988 when there is more than one recipient in a message. However, when the
5989 verification involves a callout, it is possible for different callout
5990 options to be set for different recipients. It is too complicated to keep
5991 track of this in the cache, so now Exim always runs a verification when a
5992 callout is required, relying on the callout cache for the optimization.
5993 The overhead is duplication of the address routing, but this should not be
5996 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5997 to be verified with callout=postmaster, and the main callout worked but the
5998 postmaster check failed, the verification correctly failed. However, if a
5999 subsequent RCPT command asked for sender verification *without* the
6000 postmaster check, incorrect caching caused this verification also to fail,
6003 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6004 it was not caching the DNS options (qualify_single, search_parents) that
6005 were used when the lookup failed. A subsequent lookup with different
6006 options therefore always gave the same answer, though there were cases
6007 where it should not have. (Example: a "domains = !$mx_any" option on a
6008 dnslookup router: the "domains" option is always processed without any
6009 widening, but the router might have qualify_single set.) Now Exim uses the
6010 cached value only when the same options are set.
6012 32. Added John Jetmore's "exipick" utility to the distribution.
6014 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6015 than a timeout (e.g. a certificate is required, and is not provided), an
6016 Exim server now closes the connection immediately. Previously it waited for
6017 the client to close - but if the client is SSL, it seems that they each
6018 wait for each other, leading to a delay before one of them times out.
6020 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6021 maintained 0.8.x compatibility because I don't think many are using it, and
6022 it is clearly obsolete.
6024 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6027 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6028 client certificate was expired. A simple patch fixes this, though I don't
6029 understand the full logic of why the verify callback is called multiple
6032 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6033 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6034 which causes problems with some clients (such as the Certicom SSL Plus
6035 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6036 disables the coutermeasure allowing Eudora to connect."
6038 38. Exim was not checking that a write() to a log file succeeded. This could
6039 lead to Bad Things if a log got too big, in particular if it hit a file
6040 size limit. Exim now panics and dies if it cannot write to a log file, just
6041 as it does if it cannot open a log file.
6043 39. Modified OS/Makefile-Linux so that it now contains
6045 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6047 The two -D definitions ensure that Exim is compiled with large file
6048 support, which makes it possible to handle log files that are bigger than
6051 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6052 instance) a domain was checked against a named list that involved a lookup,
6053 causing $domain_data to be set, then another domain was checked against the
6054 same list, then the first domain was re-checked, the value of $domain_data
6055 after the final check could be wrong. In particular, if the second check
6056 failed, it could be set empty. This bug probably also applied to
6059 41. The strip_trailing_dot option was not being applied to the address given
6060 with the -f command-line option.
6062 42. The code for reading a message's header from the spool was incrementing
6063 $received_count, but never initializing it. This meant that the value was
6064 incorrect (doubled) while delivering a message in the same process in which
6065 it was received. In the most common configuration of Exim, this never
6066 happens - a fresh exec is done - but it can happen when
6067 deliver_drop_privilege is set.
6069 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6070 it now includes up to 150 characters of the unexpected data in the log
6073 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6074 and building data strings. The size of both of these buffers was 10 000
6075 bytes - far larger than anybody would *ever* want, thought I. Needless to
6076 say, somebody hit the limit. I have increased the maximum line length to
6077 20 000 and the maximum data length of concatenated lines to 100 000. I have
6078 also fixed two bugs, because there was no checking on these buffers. Tsk,
6079 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6080 buffer is too small.
6082 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6083 lsearch lookups. Now it does.
6085 46. When parsing a route_list item in a manualroute router, a fixed-length
6086 buffer was used for the list of hosts. I made this 1024 bytes long,
6087 thinking that nobody would ever have a list of hosts that long. Wrong.
6088 Somebody had a whole pile of complicated expansion conditions, and the
6089 string was silently truncated, leading to an expansion error. It turns out
6090 that it is easier to change to an unlimited length (owing to other changes
6091 that have happened since this code was originally written) than to build
6092 structure for giving a limitation error. The length of the item that
6093 expands into the list of hosts is now unlimited.
6095 47. The lsearch lookup could not handle data where the length of text line was
6096 more than 4095 characters. Such lines were truncated, leading to shortened
6097 data being returned. It should now handle lines of any length.
6099 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6100 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6103 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6104 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6105 no longer happens in comment lines. A list of replacements is now placed
6106 at the head of all of the source files, except those whose only change is
6107 to replace PERL_COMMAND in the very first #! line.
6109 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6110 messages on the queue, with a bottom-up merge sort, using code contributed
6111 by Michael Haardt. This should make operations like -bp somewhat faster on
6112 large queues. It won't affect queue runners, except when queue_run_in_order
6115 51. Installed eximstats 1.31 in the distribution.
6117 52. Added support for SRV lookups to the dnslookup router.
6119 53. If an ACL referred to $message_body or $message_body_end, the value was not
6120 reset for any messages that followed in the same SMTP session.
6122 54. The store-handling optimization for building very long strings was not
6123 differentiating between the different store pools. I don't think this
6124 actually made any difference in practice, but I've tidied it.
6126 55. While running the routers to verify a sender address, $sender_address
6127 was still set to the sender address. This is wrong, because when routing to
6128 send a bounce to the sender, it would be empty. Therefore, I have changed
6129 it so that, while verifying a sender address, $sender_address is set to <>.
6130 (There is no change to what happens when verifying a recipient address.)
6132 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6133 target A or AAAA records (if not already returned) without resetting the
6134 qualify_single or search_parents options of the DNS resolver. These are
6135 inappropriate in this case because the targets of MX and SRV records must
6136 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6137 target that, when qualified, matched something in the local domain. These
6138 two options are now turned off when doing these lookups.
6140 57. It seems that at least some releases of Reiserfs (which does not have the
6141 concept of a fixed number of inodes) returns zero and not -1 for the
6142 number of available inodes. This interacted badly with check_spool_inodes,
6143 which assumed that -1 was the "no such thing" setting. What I have done is
6144 to check that the total number of inodes is greater than zero before doing
6145 the test of how many are available.
6147 58. When a "warn" ACL statement has a log_message modifier, the message is
6148 remembered, and not repeated. This is to avoid a lot of repetition when a
6149 message has many recipients that cause the same warning to be written.
6150 However, Exim was preserving the list of already written lines for an
6151 entire SMTP session, which doesn't seem right. The memory is now reset if a
6152 new message is started.
6154 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6155 debugging output unless log_rewrite was also set.
6157 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6158 of (int)(handle) when we know that handle contains (void *)(-1).
6160 61. The Exim daemon panic-logs an error return when it closes the incoming
6161 connection. However "connection reset by peer" seems to be common, and
6162 isn't really an error worthy of noting specially, so that particular error
6165 62. When Exim is trying to find all the local interfaces, it used to panic and
6166 die if the ioctl to get the interface flags failed. However, it seems that
6167 on at least one OS (Solaris 9) it is possible to have an interface that is
6168 included in the list of interfaces, but for which you get a failure error
6169 for this call. This happens when the interface is not "plumbed" into a
6170 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6171 failure of the "get flags" call assumes that the interface is down.
6173 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6174 makes life easier for people who are doing arithmetic on fields extracted
6175 from dates, where you often get leading zeros that should not be
6176 interpreted as octal.
6178 64. Added qualify_domain to the redirect router, to override the global
6181 65. If a pathologically long header line contained very many addresses (the
6182 report of this problem mentioned 10 000) and each of them was rewritten,
6183 Exim could use up a very large amount of memory. (It kept on making new
6184 copies of the header line as it rewrote, and never released the old ones.)
6185 At the expense of a bit more processing, the header rewriting function has
6186 been changed so that it no longer eats memory in this way.
6188 66. The generation of the Received: header has been moved from the time that a
6189 message starts to be received, to the time that it finishes. The timestamp
6190 in the Received: header should now be very close to that of the <= log
6191 line. There are two side-effects of this change:
6193 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6194 logged header lines no longer include the local Received: line, because
6195 it has not yet been created. The same applies to a copy of the message
6196 that is returned to a non-SMTP sender when a message is rejected.
6198 (b) When a filter file is tested using -bf, no additional Received: header
6199 is added to the test message. After some thought, I decided that this
6202 This change does not affect the value of $received_for. It is still set
6203 after address rewriting, but before local_scan() is called.
6205 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6207 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6208 gave an unhelpful panic error message, and a defer error. I have managed to
6209 change this behaviour so that it now rejects any supplied certificate,
6210 which seems right, as the list of acceptable certificates is empty.
6212 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6213 gave an unhelpful defer error. I have not managed to make this reject any
6214 supplied certificates, but the error message it gives is "no certificate
6215 supplied", which is not helpful.
6217 70. exigrep's output now also includes lines that are not associated with any
6218 message, but which match the given pattern. Implemented by a patch from
6219 Martin Sluka, which also tidied up the Perl a bit.
6221 71. Recipient callout verification, like sender verification, was using <> in
6222 the MAIL FROM command. This isn't really the right thing, since the actual
6223 sender may affect whether the remote host accepts the recipient or not. I
6224 have changed it to use the actual sender in the callout; this means that
6225 the cache record is now keyed on a recipient/sender pair, not just the
6226 recipient address. There doesn't seem to be a real danger of callout loops,
6227 since a callout by the remote host to check the sender would use <>.
6228 [SEE ABOVE: changed after hitting problems.]
6230 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6231 temporary errors. However, in the case of such a code being given after
6232 the end of a data transmission (i.e. after ".") Exim was failing to write
6233 a retry record for the message. (Yes, there was some broken host that was
6234 actually sending 8xx at this point.)
6236 73. An unknown lookup type in a host list could cause Exim to panic-die when
6237 the list was checked. (An example that provoked this was putting <; in the
6238 middle of a list instead of at the start.) If this happened during a DATA
6239 ACL check, a -D file could be left lying around. This kind of configuration
6240 error no longer causes Exim to die; instead it causes a defer error. The
6241 incident is still logged to the main and panic logs.
6243 74. Buglet left over from Exim 3 conversion. The message "too many messages
6244 in one connection" was written to the rejectlog but not the mainlog, except
6245 when address rewriting (yes!) was being logged.
6247 75. Added write_rejectlog option.
6249 76. When a system filter was run not as root (that is, when system_filter_user
6250 was set), the values of the $n variables were not being returned to the
6251 main process; thus, they were not subsequently available in the $sn
6254 77. Added +return_path_on_delivery log selector.
6256 78. A connection timeout was being treated differently from recipients deferred
6257 when testing hosts_max_try with a message that was older than the host's
6258 retry timeout. (The host should not be counted, thus allowing all hosts to
6259 be tried at least once before bouncing.) This may have been the cause of an
6260 occasionally reported bug whereby a message would remain on the queue
6261 longer than the retry timeout, but would be bounced if a delivery was
6262 forced. I say "may" because I never totally pinned down the problem;
6263 setting up timeout/retry tests is difficult. See also the next item.
6265 79. The ultimate address timeout was not being applied to errors that involved
6266 a combination of host plus message (for example, a timeout on a MAIL
6267 command). When an address resolved to a number of possible hosts, and they
6268 were not all tried for each delivery (e.g. because of hosts_max_try), a
6269 message could remain on the queue longer than the retry timeout.
6271 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6274 81. Fixed an obscure SMTP outgoing bug which required at least the following
6275 conditions: (a) there was another message waiting for the same server;
6276 (b) the server returned 5xx to all RCPT commands in the first message so
6277 that the message was not completed; (c) the server dropped the connection
6278 or gave a negative response to the RSET that Exim sends to abort the
6279 transaction. The observed case was a dropped connection after DATA that had
6280 been sent in pipelining mode. That is, the server had advertised PIPELINING
6281 but was not implementing it correctly. The effect of the bug was incorrect
6282 behaviour, such as trying another host, and this could lead to a crash.
6288 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6289 and daemon.c were passed as pointers to ints; they should have been
6290 pointers to socklen_t variables (which are typically unsigned ints).
6292 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6295 3. Fixed a really odd bug that affected only the testing scheme; patching a
6296 certain fixed string in the binary changed the value of another string that
6297 happened to be identical to the end of the original first string.
6299 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6300 name", it returns that address as the IP address. On some operating
6301 systems (e.g. Solaris), it also passes back the IP address string as the
6302 "host name". However, on others (e.g. Linux), it passes back an empty
6303 string. Exim wasn't checking for this, and was changing the host name to an
6304 empty string, assuming it had been canonicalized.
6306 5. Although rare, it is permitted to have more than one PTR record for a given
6307 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6308 all the names associated with an address, because they do in Solaris.
6309 However, it seems that they do not in Linux for data that comes from the
6310 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6311 I found this out when I moved to a new Linux workstation and tried to run
6312 the Exim test suite.
6314 To get round this problem I have changed the code so that it now does its
6315 own call to the DNS to look up PTR records when searching for a host name.
6316 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6317 addresses that are only in /etc/hosts are still found.
6319 This behaviour is, however, controlled by an option called host_lookup_
6320 order, which defaults to "bydns:byaddr". If people want to use the other
6321 order, or indeed, just use one or the other means of lookup, they can
6322 specify it in this variable.
6324 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6325 some operating systems, this comes back from gethostbyaddr() as an empty
6326 string, and this is what Exim used to test for. However, it seems that in
6327 other systems, "." is yielded. Exim now tests for this case too.
6329 7. The values of check_spool_space and check_log_space are now held internally
6330 as a number of kilobytes instead of an absolute number of bytes. If a
6331 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6332 kilobyte. This means that much larger values can be stored.
6334 8. Exim monitor: an attempt to get the action menu when not actually pointing
6335 at a message produces an empty menu entitled "No message selected". This
6336 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6337 no entries in it ("Shell widget menu has zero width and/or height"). So I
6338 have added a single, blank menu entry in this case.
6340 9. Added ${quote_local_part.
6342 10. MIME decoding is now applied to the contents of Subject: header lines when
6345 11. Now that a reference to $sender_host_address automatically causes a reverse
6346 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6347 host lookup before query-style lookups in lists that might use this
6348 variable. This has therefore been abolished, and the "net-" prefix is no
6349 longer necessary for query-style lookups.
6351 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6352 have been a typo for LFLAGS, so it has been changed.
6354 13. The install script calls Exim with "-C /dev/null" in order to find the
6355 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6356 to be output. However, since Exim outputs its version number before the
6357 error, it didn't break the script. It just looked ugly. I fixed this by
6358 always allowing "-C /dev/null" if the caller is root.
6360 14. Ignore overlarge ACL variable number when reading spool file - insurance
6361 against a later release with more variables having written the file.
6363 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6364 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6365 this, as well as the form without the "IPv6" on the front.
6367 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6368 OS/Makefile-Darwin file.
6370 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6371 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6373 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6374 variable. However, some OS use ints for inodes. I've added cast to long int
6375 to get rid of the compiler warning.
6377 19. I had forgotten to lock out "/../" in configuration file names when
6378 ALT_CONFIG_PREFIX was set.
6380 20. Routers used for verification do not need to specify transports. However,
6381 if such a router generated a host list, and callout was configured, Exim
6382 crashed, because it could not find a port number from the (non-existent)
6383 transport. It now assumes port 25 in this circumstance.
6385 21. Added the -t option to exigrep.
6387 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6388 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6389 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6390 them is set, LOOKUP_LSEARCH is forced.
6392 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6394 24. Added sender and host information to the "rejected by local_scan()" log
6395 line; previously there was no indication of these.
6397 25. Added .include_if_exists.
6399 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6400 Linux. It turns out that not all file systems support this. Apparently some
6401 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6402 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6403 not supported on the file descriptor, is now ignored when Exim is trying to
6404 sync a directory. This applies only to Linux.
6406 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6408 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6409 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6412 29. Added support for maildirsize files from supplied patch (modified a bit).
6414 30. The use of :fail: followed by an empty string could lead Exim to respond to
6415 sender verification failures with (e.g.):
6417 550 Verification failed for <xxx>
6418 550 Sender verify failed
6420 where the first response line was missing the '-' that indicates it is not
6421 the final line of the response.
6423 31. The loop for finding the name of the user that called Exim had a hardwired
6424 limit of 10; it now uses the value of finduser_retries, which is used for
6425 all other user lookups.
6427 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6430 33. Exim was neglecting to zero errno before one call of strtol() when
6431 expanding a string and expecting an integer value. On some systems this
6432 resulted in spurious "integer overflow" errors. Also, it was casting the
6433 result into an int without checking.
6435 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6436 did not work. The code looks as if it has *never* worked, though it appears
6437 to have been documented since at least release 1.62. I have made it work.
6439 35. The "timeout_DNS" error in retry rules, also documented since at least
6440 1.62, also never worked. As it isn't clear exactly what this means, and
6441 clearly it isn't a major issue, I have abolished the feature by treating it
6442 as "timeout", and writing a warning to the main and panic logs.
6444 36. The display of retry rules for -brt wasn't always showing the error code
6447 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6448 timeout_connect_A, timeout_connect_MX.
6450 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6451 to the empty sender.
6453 39. The daemon was not analysing the content of -oX till after it had closed
6454 stderr and disconnected from the controlling terminal. This meant that any
6455 syntax errors were only noted on the panic log, and the return code from
6456 the command was 0. By re-arranging the code a little, I've made the
6457 decoding happen first, so such errors now appear on stderr, and the return
6458 code is 1. However, the actual setting up of the sockets still happens in
6459 the disconnected process, so errors there are still only recorded on the
6462 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6463 connections (as happens on some IP stacks) was logged at start up time as
6464 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6465 it from "IPv6 and IPv4", which means that two separate sockets are being
6468 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6469 says whether AF_INET or AF_INET6 was passed as an argument.
6471 42. Exiwhat output was messed up when time zones were included in log
6474 43. Exiwhat now gives more information about the daemon's listening ports,
6475 and whether -tls-on-connect was used.
6477 44. The "port" option of the smtp transport is now expanded.
6479 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6480 silently ignored. Now an error message is written to the main and panic
6483 46. There's a new ACL modifier called "logwrite" which writes to a log file
6484 as soon as it is encountered.
6486 47. Added $local_user_uid and $local_user_gid at routing time.
6488 48. Exim crashed when trying to verify a sender address that was being
6491 49. Exim was recognizing only a space character after ".include". It now also
6492 recognizes a tab character.
6494 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6495 extracting the relevant information from the specification. The man page no
6496 longer contains scrambled data for the -d option, and I've added a section
6497 at the front about calling Exim under different names.
6499 51. Added "extra_headers" argument to the "mail" command in filter files.
6501 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6504 53. Installed eximstats 1.29.
6506 54. Added transport_filter_timeout as a generic transport option.
6508 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6509 Cc: header lines. This was required by RFC 822, but it not required by RFC
6512 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6513 incoming messages that did not have them. Now it does so only if the
6514 message originates locally, that is, if there is no associated remote host
6515 address. When Resent- header lines are present, this applies to the Resent-
6516 lines rather than the non-Resent- lines.
6518 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6519 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6521 58. Messages for configuration errors now include the name of the main
6522 configuration file - useful now that there may be more than one file in a
6523 list (.included file names were always shown).
6525 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6526 for those rare installations that do not start the daemon as root or run it
6527 setuid root. I've cut out the call to initgroups() if the daemon is not
6530 60. The Exim user and group can now be bound into the binary as text strings
6531 that are looked up at the start of Exim's processing.
6533 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6535 62. Added $mailstore_basename variable.
6537 63. Installed patch to sieve.c from Michael Haardt.
6539 64. When Exim failed to open the panic log after failing to open the main log,
6540 the original message it was trying to log was written to stderr and debug
6541 output, but if they were not available (the usual case in production), it
6542 was lost. Now it is written to syslog before the two lines that record the
6543 failures to open the logs.
6545 65. Users' Exim filters run in subprocesses under the user's uid. It is
6546 possible for a "deliver" command or an alias in a "personal" command to
6547 provoke an address rewrite. If logging of address rewriting is configured,
6548 this fails because the process is not running as root or exim. There may be
6549 a better way of dealing with this, but for the moment (because 4.30 needs
6550 to be released), I have disabled address rewrite logging when running a
6551 filter in a non-root, non-exim process.
6557 1. The buildconfig auxiliary program wasn't quoting the value set for
6558 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6559 not defined. This bug was masked in 4.22 by the effect that was fixed in
6562 2. Some messages that were rejected after a message id was allocated were
6563 shown as "incomplete" by exigrep. It no longer does this for messages that
6564 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6566 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6567 have allow_domain_literals set, the ID did not get logged in the <= line.
6568 Domain literals are now always recognized in Message-ID: header lines.
6570 4. The first argument for a ${extract expansion item is the key name or field
6571 number. Leading and trailing spaces in this item were not being ignored,
6572 causing some misleading effects.
6574 5. When deliver_drop_privilege was set, single queue runner processes started
6575 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6576 same command in the process it spins off) were not dropping privilege.
6578 6. When the daemon running as "exim" started a queue runner, it always
6579 re-executed Exim in the spun-off process. This is a waste of effort when
6580 deliver_drop_privilege is set. The new process now just calls the
6581 queue-runner function directly.
6587 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6590 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6591 ignored. Though the use of -oP was forcing the writing of a pid file, it
6592 was always written to the default place.
6594 3. If the message "no IP address found for host xxxx" is generated during
6595 incoming verification, it is now followed by identification of the incoming
6596 connection (so you can more easily find what provoked it).
6598 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6600 5. Added some features to "harden" Exim a bit more against certain attacks:
6602 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6603 be put in Local/Makefile. This is like the never_users runtime option,
6604 but it cannot be overridden. The default setting is "root".
6606 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6607 prefix string with which any file named in a -C command line option
6610 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6611 is retained for -C and -D only if the caller of Exim is root. Without
6612 it, the exim user may also use -C and -D and retain privilege.
6614 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6615 command line option is disabled.
6617 6. Macro names set by the -D option must start with an upper case letter, just
6618 like macro names defined in the configuration file.
6620 7. Added "dereference=" facility to LDAP.
6622 8. Two instances of the typo "uknown" in the source files are fixed.
6624 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6625 the Configure-Makefile script screwed up while processing it.
6627 10. Incorporated PCRE 4.4.
6629 11. The SMTP synchronization check was not operating right at the start of an
6630 SMTP session. For example, it could not catch a HELO sent before the client
6631 waited for the greeting. There is now a check for outstanding input at the
6632 point when the greeting is written. Because of the duplex, asynchronous
6633 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6634 way, but not yet received, when the check is performed.
6636 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6637 on TCP/IP sockets, because this apparently causes some broken clients to
6640 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6641 unchanged) from the Cygwin maintainer.
6643 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6644 was supported instead of testing for mbx. Effectively a typo.
6646 15. The spa authenticator server code was not checking that the input it
6647 received was valid base64.
6649 16. The debug output line for the "set" modifier in ACLs was not showing the
6650 name of the variable that was being set.
6652 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6654 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6655 lookup on its own. Something else was needed to trigger the lookup. For
6656 example, a match in host_lookup or the need for a host name in a host list.
6657 Now, if $sender_host_name is referenced and the host name has not yet been
6658 looked up, a lookup is performed. If the lookup fails, the variable remains
6659 empty, and $host_lookup_failed is set to "1".
6661 19. Added "eqi" as a case-independent comparison operator.
6663 20. The saslauthd authentication condition could segfault if neither service
6664 nor realm was specified.
6666 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6667 error message that was logged was misleading, and incoming SMTP
6668 connections were dropped. The message is now more accurate, and temporary
6669 errors are given to SMTP connections.
6671 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6672 (except RSET) with a 421 error, until QUIT is received. However, it was
6673 failing to send a response to QUIT.
6675 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6676 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6677 ACL is now run after the helo_try_verify_hosts code.
6679 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6680 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6681 case-independent, so other case variants are also recognized). Apparently
6682 some systems use these upper case variants.
6684 25. If more than two messages were waiting for the same host, and a transport
6685 filter was specified for the transport, Exim sent two messages over the
6686 same TCP/IP connection, and then failed with "socket operation on non-
6687 socket" when it tried to send the third.
6689 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6691 27. The extern definition of crypt16() in expand.c was not being excluded when
6692 the OS had its own crypt16() function.
6694 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6695 as a preferred synonym for return_size_limit, both as an option and as an
6698 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6700 30. Changed the default configuration ACL to relax the local part checking rule
6701 for addresses that are not in any local domains. For these addresses,
6702 slashes and pipe symbols are allowed within local parts, but the sequence
6703 /../ is explicitly forbidden.
6705 31. SPA server authentication was not clearing the challenge buffer before
6708 32. log_message in a "warn" ACL statement was writing to the reject log as
6709 well as to the main log, which contradicts the documentation and doesn't
6710 seem right (because no rejection is happening). So I have stopped it.
6712 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6713 However, I am unable to do any testing of this.
6715 34. Fixed an infelicity in the appendfile transport. When checking directories
6716 for a mailbox, to see if any needed to be created, it was accidentally
6717 using path names with one or more superfluous leading slashes; tracing
6718 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6720 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6721 contents are added to the log line that is written for every discarded
6722 recipient. (Previously a log_message setting was ignored.)
6724 36. The ${quote: operator now quotes the string if it is empty.
6726 37. The install script runs exim in order to find its version number. If for
6727 some reason other than non-existence or emptiness, which it checks, it
6728 could not run './exim', it was installing it with an empty version number,
6729 i.e. as "exim-". This error state is now caught, and the installation is
6732 38. An argument was missing from the function that creates an error message
6733 when Exim fails to connect to the socket for saslauthd authentication.
6734 This could cause Exim to crash, or give a corrupted message.
6736 39. Added isip, isip4, and isip6 to ${if conditions.
6738 40. The ACL variables $acl_xx are now saved with the message, and can be
6739 accessed later in routers, transports, and filters.
6741 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6742 strings in the file are not string-expanded.
6744 42. If a MAIL command specified a SIZE value that was too large to fit into an
6745 int variable, the check against message_size_limit failed. Such values are
6746 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6747 day this will have to be increased, but I don't think I want to be around
6748 when emails are that large.
6755 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6756 iconv() is not standard in FreeBSD.
6758 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6759 IPv6 enabled. The observed symptom was a segmentation fault on return from
6760 the function os_common_find_running_interfaces() in src/os.c.
6762 3. In the check_special_case() function in daemon.c I had used "errno" as an
6763 argument name, which causes warnings on some systems. This was basically a
6764 typo, since it was named "eno" in the comments!
6766 4. The code that waits for the clock to tick (at a resolution of some fraction
6767 of a second) so as to ensure message-id uniqueness was always waiting for
6768 at least one whole tick, when it could have waited for less. [This is
6769 almost certainly not relevant at current processor speeds, where it is
6770 unlikely to ever wait at all. But we try to future-proof.]
6772 5. The function that sleeps for a time interval that includes fractions of a
6773 second contained a race. It did not block SIGALRM between setting the
6774 timer, and suspending (a couple of lines later). If the interval was short
6775 and the sigsuspend() was delayed until after it had expired, the suspension
6776 never ended. On busy systems this could lead to processes getting stuck for
6779 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6780 process, before it forks any delivery processes. The open lookup caching
6781 mechanism meant that the open file or database connection was passed into
6782 the delivery process. The problem was that delivery processes always tidy
6783 up cached lookup data. This could cause a problem for the next delivery
6784 process started by the queue runner, because the external queue runner
6785 process does not know about the closure. So the next delivery process
6786 still has data in the lookup cache. In the case of a file lookup, there was
6787 no problem because closing a file descriptor in a subprocess doesn't affect
6788 the parent. However, if the lookup was caching a connection to a database,
6789 the connection was closed, and the second delivery process was likely to
6790 see errors such as "PGSQL: query failed: server closed the connection
6791 unexpectedly". The problem has been fixed by closing all cached lookups
6792 in a queue runner before running a delivery process.
6794 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6795 seem to have the "const" qualifier which it has on other OS. I've
6798 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6799 *of the same type* (client or server) with the same public name that an
6800 error should be diagnosed.
6802 9. When Exim looked up a host name for an IP address, but failed to find the
6803 original IP address when looking up the host name (a safety check), it
6804 output the message "<ip address> does not match any IP for NULL", which was
6805 confusing, to say the least. The bug was that the host name should have
6806 appeared instead of "NULL".
6808 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6809 user that is built into the binary, and the -C or -D options is used, root
6810 privilege is dropped before the configuration file is read. In addition,
6811 logging is switched to stderr instead of the normal log files. If the
6812 configuration then re-defines the Exim user, the unprivileged environment
6813 is probably not what is expected, so Exim logs a panic warning message (but
6816 However, if deliver_drop_privilege is set, the unprivileged state may well
6817 be exactly what is intended, so the warning has been cut out in that case,
6818 and Exim is allowed to try to write to its normal log files.
6824 1. smtp_return_error_details was not giving details for temporary sender
6825 or receiver verification errors.
6827 2. Diagnose a configuration error if two authenticators have the same public
6830 3. Exim used not to create the message log file for a message until the first
6831 delivery attempt. This could be confusing when incoming messages were held
6832 for policy or load reasons. The message log file is now created at the time
6833 the message is received, and an initial "Received" line is written to it.
6835 4. The automatically generated man page for command line options had a minor
6836 bug that caused no ill effects; however, a more serious problem was that
6837 the procedure for building the man page automatically didn't always
6838 operate. Consequently, release 4.20 contains an out-of-date version. This
6839 shouldn't happen again.
6841 5. When building Exim with embedded Perl support, the script that builds the
6842 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6843 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6845 6. The freeze_tell option was not being used for messages that were frozen on
6846 arrival, either by an ACL or by local_scan().
6848 7. Added the smtp_incomplete_transaction log selector.
6850 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6851 was accepting AUTH without a new EHLO.
6853 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6854 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6857 10. Logging of TCP/IP connections (when configured) now happens in the main
6858 daemon process instead of the child process, so that the TCP/IP connection
6859 count is more accurate (but it can never be perfect).
6861 11. The use of "drop" in a nested ACL was not being handled correctly in the
6862 outer ACL. Now, if condition failure induced by the nested "drop" causes
6863 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6864 or "require"), the connection is dropped.
6866 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6867 that yield "discard" can now be used with an "accept" or a "discard" verb,
6868 but an error is generated for any others (because I can't see a useful way
6869 to define what should happen).
6871 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6872 are now processed in the same way as lines in the Exim configuration file.
6873 In particular, continuation lines are supported.
6875 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6877 15. Added -ti meaning -t -i.
6879 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6880 domains, and warn by logging if others are found.
6882 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6883 returned from GIFCONF when Exim is trying to find the list of interfaces on
6884 a host. The code in os.c has been modified to copy each ifreq to an aligned
6885 structure in all cases.
6887 Also, in some cases, the returned ifreq's were being copied to a 'struct
6888 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6889 means the last couple of bytes of an IPv6 address could be chopped if the
6890 ifreq contained only a normal sockaddr (14 bytes storage).
6892 18. Named domain lists were not supported in the hosts_treat_as_local option.
6893 An entry such as +xxxx was not recognized, and was treated as a literal
6896 19. Ensure that header lines added by a DATA ACL are included in the reject log
6897 if the ACL subsequently rejects the message.
6899 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6900 MD5 (which is deprecated).
6902 21. When testing a filter file using -bf, Exim was writing a message when it
6903 took the sender from a "From " line in the message, but it was not doing so
6904 when it took $return_path from a Return-Path: header line. It now does.
6906 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6907 with a valid header line field name (a series of printing characters
6908 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6910 23. Changed "disc" in the source to "disk" to conform to the documentation and
6911 the book and for uniformity.
6913 24. Ignore Sendmail's -Ooption=value command line item.
6915 25. When execve() failed while trying to run a command in a pipe transport,
6916 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6917 could be confused with a return value of 69 from the command itself. This
6918 has been changed to 127, the value the shell returns if it is asked to run
6919 a non-existent command. The wording for the related log line suggests a
6920 non-existent command as the problem.
6922 26. If received_header_text expands to an empty string, do not add a Received:
6923 header line to the message. (Well, it adds a token one on the spool, but
6924 marks it "old" so that it doesn't get used or transmitted.)
6926 27. Installed eximstats 1.28 (addition of -nt option).
6928 28. There was no check for failure on the call to getsockname() in the daemon
6929 code. This can fail if there is a shortage of resources on the system, with
6930 ENOMEM, for example. A temporary error is now given on failure.
6932 29. Contrary to the C standard, it seems that in some environments, the
6933 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6934 program. Exim now does this explicitly; it affects the formatting of
6935 timestamps using strftime().
6937 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6938 complaints. I've now initialized all the variables, to avoid this.
6940 32. Header lines added by a system filter were not being "seen" during
6941 transport-time rewrites.
6943 33. The info_callback() function passed to OpenSSL is set up with type void
6944 (*)(SSL *, int, int), as described somewhere. However, when calling the
6945 function (actually a macro) that sets it up, the type void(*)() is
6946 expected. I've put in a cast to prevent warnings from picky compilers.
6948 34. If a DNS black list lookup found a CNAME record, but there were no A
6949 records associated with the domain it pointed at, Exim crashed.
6951 35. If a DNS black list lookup returned more than one A record, Exim ignored
6952 all but the first. It now scans all returned addresses if a particular IP
6953 value is being sought. In this situation, the contents of the
6954 $dnslist_value variable are a list of all the addresses, separated by a
6957 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6958 to accept a host name and all its aliases if the forward lookup for any of
6959 them yielded the IP address of the incoming connection. Now it accepts only
6960 those names whose forward lookup yields the correct IP address. Any other
6961 names are discarded. This closes a loophole whereby a rogue DNS
6962 administrator could create reverse DNS records to break through a
6963 wildcarded host restriction in an ACL.
6965 37. If a user filter or a system filter that ran in a subprocess used any of
6966 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6967 the wrong values were passed to the pipe command ($thisaddress had the
6968 value of $0, $0 had the value of $1, etc). This bug was introduced by
6969 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6971 38. Improved the line breaking for long SMTP error messages from ACLs.
6972 Previously, if there was no break point between 40 and 75 characters, Exim
6973 left the rest of the message alone. Two changes have been made: (a) I've
6974 reduced the minimum length to 35 characters; (b) if it can't find a break
6975 point between 35 and 75 characters, it looks ahead and uses the first one
6976 that it finds. This may give the occasional overlong line, but at least the
6977 remaining text gets split now.
6979 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6980 file descriptors might be low, and that setting 1000 would always raise it.
6981 It turns out that in some environments, the limit is already over 1000 and
6982 that lowering it causes trouble. So now Exim takes care not to decrease it.
6984 40. When delivering a message, the value of $return_path is set to $sender_
6985 address at the start of routing (routers may change the value). By an
6986 oversight, this default was not being set up when an address was tested by
6987 -bt or -bv, which affected the outcome if any router or filter referred to
6990 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6991 log only when "message" or "log_message" are set. However, if one of the
6992 conditions was an address verification, or a call to a nested ACL, the
6993 messages generated by the underlying test were being passed through. This
6994 no longer happens. The underlying message is available in $acl_verify_
6995 message for both "message" and "log_message" expansions, so it can be
6996 passed through if needed.
6998 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6999 new expansion $bh_ to give the encoded byte string without charset
7000 translation. Translation happens only if iconv() is available; HAVE_ICONV
7001 indicates this at build time. HEADERS_CHARSET gives the charset to
7002 translate to; headers_charset can change it in the configuration, and
7003 "headers charset" can change it in an individual filter file.
7005 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7006 that creates RFC 2047 encoded "words" labels them as that charset instead
7007 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7008 expansion operator; (ii) when Exim creates a From: line for a local
7009 message; (iii) when a header line is rewritten to include a "phrase" part.
7011 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7012 buggy, causing it to skip the first lines of messages whose message ID
7013 ended in 'D'. This would not have bitten before Exim release 4.14, because
7014 message IDs were unlikely to end in 'D' before then. The effect was to have
7015 incorrect size information for certain domains.
7017 45. #include "config.h" was missing at the start of the crypt16.c module. This
7018 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7021 46. If there was a timeout during a "random" callout check, Exim treated it as
7022 a failure of the random address, and carried on sending RSET and the real
7023 address. If the delay was just some slowness somewhere, the response to the
7024 original RCPT would be taken as a response to RSET and so on, causing
7025 mayhem of various kinds.
7027 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7028 when I implemented it. It didn't allow for the fact that some option values
7029 may legitimately be negative (e.g. size_addition), and it didn't even do
7030 the right test for positive values.
7032 48. Domain names in DNS records are case-independent. Exim always looks them up
7033 in lower case. Some resolvers return domain names in exactly the case they
7034 appear in the zone file, that is, they may contain uppercase letters. Not
7035 all resolvers do this - some return always lower case. Exim was treating a
7036 change of case by a resolver as a change of domain, similar to a widening
7037 of a domain abbreviation. This triggered its re-routing code and so it was
7038 trying to route what was effectively the same domain again. This normally
7039 caused routing to fail (because the router wouldn't handle the domain
7040 twice). Now Exim checks for this case specially, and just changes the
7041 casing of the domain that it ultimately uses when it transmits the message
7044 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7047 50. If a filter generated a file delivery with a non-absolute name (possible if
7048 no home directory exists for the router), the forbid_file option was not
7051 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7052 the existing equality matching.
7054 52. Exim was using ints instead of ino_t variables in some places where it was
7055 dealing with inode numbers.
7057 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7058 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7059 called TMPDIR, and if it finds it is different, it changes its value.
7061 54. The smtp_printf() function is now made available to local_scan() so
7062 additional output lines can be written before returning. There is also an
7063 smtp_fflush() function to enable the detection of a dropped connection.
7064 The variables smtp_input and smtp_batched_input are exported to
7067 55. Changed the default runtime configuration: the message "Unknown user"
7068 has been removed from the ACL, and instead placed on the localuser router,
7069 using the cannot_route_message feature. This means that any verification
7070 failures that generate their own messages won't get overridden. Similarly,
7071 the "Unrouteable address" message that was in the ACL for unverifiable
7072 relay addresses has also been removed.
7074 56. Added hosts_avoid_esmtp to the smtp transport.
7076 57. The exicyclog script was not checking for the esoteric option
7077 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7078 will work only if exicyclog is run under the appropriate euid.
7080 58. Following a discussion on the list, the rules by which Exim recognises line
7081 endings on incoming messages have been changed. The -dropcr and drop_cr
7082 options are now no-ops, retained only for backwards compatibility. The
7083 following line terminators are recognized: LF CRLF CR. However, special
7084 processing applies to CR:
7086 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7087 nor a local message in the state where . is a terminator.
7089 (ii) If a bare CR is encountered in a header line, an extra space is added
7090 after the line terminator so as not to end the header. The reasoning
7091 behind this is that bare CRs in header lines are most likely either
7092 to be mistakes, or people trying to play silly games.
7094 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7095 was being incorrectly given as 18 bytes larger than it should have been.
7096 This is a VOB (very old bug).
7098 60. This may never have affected anything current, but just in case it has:
7099 When the local host is found other than at the start of a list of hosts,
7100 the local host, those with the same MX, and any that follow, are discarded.
7101 When the list in question was part of a longer list of hosts, the following
7102 hosts (not currently being processed) were also being discarded. This no
7103 longer happens. I'm not sure if this situation could ever has previously
7106 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7109 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7110 header. This is something that is recommended in a new Internet Draft, and
7111 is something that is documented as being done by Sendmail. There are two
7112 possible values. For messages generated by the autoreply transport, Exim
7115 Auto-Submitted: auto-replied
7117 whereas for all other generated messages (e.g. bounces) it adds
7119 Auto-Submitted: auto-generated
7121 63. The "personal" condition in filters now includes a test for the
7122 Auto-Submitted: header. If it contains the string "auto-" the message it
7123 not considered personal.
7125 64. Added rcpt_include_affixes as a generic transport option.
7127 65. Added queue_only_override (default true).
7129 66. Added the syslog_duplication option.
7131 67. If what should have been the first header line of a message consisted of
7132 a space followed by a colon, Exim was mis-interpreting it as a header line.
7133 It isn't of course - it is syntactically invalid and should therefore be
7134 treated as the start of the message body. The misbehaviour could have
7135 caused a number of strange effects, including loss of data in subsequent
7136 header lines, and spool format errors.
7138 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7139 client host had authenticated. This control can now be exercised by an ACL
7140 for more flexibility.
7142 69. By default, callouts do not happen when testing with -bh. There is now a
7143 variant, -bhc, which does actually run the callout code, including
7144 consulting and updating the callout cache.
7146 70. Added support for saslauthd authentication, courtesy of Alexander
7149 71. If statvfs() failed on the spool or log directories while checking their
7150 size for availability, Exim confusingly gave the error "space shortage".
7151 Furthermore, in debugging mode it crashed with a floating point exception.
7152 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7153 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7154 serious problem, Exim now writes to the main and panic logs when this
7155 happens, with details of the failure. It then refuses to accept the
7156 incoming message, giving the message "spool directory problem" or "log
7157 directory problem" with a 421 code for SMTP messages.
7159 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7160 0, 1, and 2 exist, because some OS complain for execs without them (see
7161 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7162 descriptors. However, the code omitted to check that the open succeeded,
7163 causing mysterious errors if for some reason the permissions on /dev/null
7164 got screwed. Now Exim writes a message to the main and panic logs, and
7165 bombs out if it can't open /dev/null.
7167 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7168 interact so that it is all more flexible. It is supposed to remain
7169 backwards compatible. Also added extra_local_interfaces.
7171 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7172 to bomb out with an assertion failure - to the client this appears as a
7173 connection drop. This problem occurs in the part of the code that was taken
7174 from the Samba project. Fortunately, the assertion is in a very simple
7175 function, so I have fixed this by reproducing the function inline in the
7176 one place where it is called, and arranging for authentication to fail
7177 instead of killing the process with assert().
7179 75. The SPA client code was not working when the server requested OEM rather
7180 than Unicode encoding.
7182 76. Added code to make require_files with a specific uid setting more usable in
7183 the case where statting the file as root fails - usually a non-root-mounted
7184 NFS file system. When this happens and the failure is EACCES, Exim now
7185 forks a subprocess and does the per-uid checking as the relevant uid.
7187 77. Added process_log_path.
7189 78. If log_file_path was not explicitly set, a setting of check_log_space or
7190 check_log_inodes was ignored.
7192 79. If a space check for the spool or log partitions fails, the incident is now
7193 logged. Of course, in the latter case the data may get lost...
7195 80. Added the %p formatting code to string_format() so that it can be used to
7196 print addresses in debug_print(). Adjusted all the address printing in the
7197 debugging in store.c to use %p rather than %d.
7199 81. There was a concern that a line of code in smtp_in.c could overflow a
7200 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7201 initially expressed, the concern was not well-founded, because trailing
7202 spaces are removed early. However, if the trailing spaces were followed by
7203 a NULL, they did not get removed, so the overflow was possible. Two fixes
7206 (a) I re-wrote the offending code in a cleaner fashion.
7207 (b) If an incoming SMTP command contains a NULL character, it is rejected
7210 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7211 runs initgroups(), so that if the Exim user is in any additional groups,
7212 they will be used during message reception.
7218 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob