1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
31 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
34 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
35 Previously this would segfault.
37 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
44 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
45 When redis_servers is set to a list of > 1 element, and the Redis servers
46 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
47 case of MOVED into a DEFER case instead, thus moving the query onto the
48 next server in the list. For a cluster of N elements, all N servers must
49 be defined in redis_servers.
51 GF/02 Catch and remove uninitialized value warning in exiqsumm
52 Check for existence of @ARGV before looking at $ARGV[0]
54 JH/01 Replace the store_release() internal interface with store_newblock(),
55 which internalises the check required to safely use the old one, plus
56 the allocate and data copy operations duplicated in both (!) of the
59 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
60 modifier. This matches the restriction on the commandline.
62 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
63 Previously only the last row was returned.
65 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
66 we assumed that tags in the header were well-formed, and parsed the
67 element content after inspecting only the first char of the tag.
68 Assumptions at that stage could crash the receive process on malformed
71 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
72 While running the DKIM ACL we operate on the Permanent memory pool so that
73 variables created with "set" persist to the DATA ACL. Also (at any time)
74 DNS lookups that fail create cache records using the Permanent pool. But
75 expansions release any allocations made on the current pool - so a dnsdb
76 lookup expansion done in the DKIM ACL releases the memory used for the
77 DNS negative-cache, and bad things result. Solution is to switch to the
78 Main pool for expansions.
79 While we're in that code, add checks on the DNS cache during store_reset,
80 active in the testsuite.
81 Problem spotted, and debugging aided, by Wolfgang Breyha.
83 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
84 When none of the hosts presented to a transport match an already-open
85 connection, close it and proceed with the list. Previously we would
86 queue the message. Spotted by Lena with Yahoo, probably involving
89 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
90 Previously a spurious "250 OK id=" response was appended to the proper
93 JH/08 The "support for" informational output now, which built with Content
94 Scanning support, has a line for the malware scanner interfaces compiled
95 in. Interface can be individually included or not at build time.
97 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
98 by the template makefile "src/EDITME". The "STREAM" support for an older
99 ClamAV interface method is removed.
101 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
102 rows affected is given instead).
104 JH/11 The runtime Berkeley DB library version is now additionally output by
105 "exim -d -bV". Previously only the compile-time version was shown.
107 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
108 SMTP connection. Previously, when one had more receipients than the
109 first, an abortive onward connection was made. Move to full support for
110 multiple onward connections in sequence, handling cutthrough connection
111 for all multi-message initiating connections.
113 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
114 routers. Previously, a multi-recipient message would fail to match the
115 onward-connection opened for the first recipient, and cause its closure.
117 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
118 a timeout on read on a GnuTLS initiating connection, resulting in the
119 initiating connection being dropped. This mattered most when the callout
120 was marked defer_ok. Fix to keep the two timeout-detection methods
123 JH/15 Relax results from ACL control request to enable cutthrough, in
124 unsupported situations, from error to silently (except under debug)
125 ignoring. This covers use with PRDR, frozen messages, queue-only and
128 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
130 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
131 metadata, resulting in a crash in free().
133 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
134 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
135 Broken also in d185889f4, with init system revamp.
137 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
138 Previously we abruptly closed the connection after reading a malware-
139 found indication; now we go on to read the "scan ok" response line,
142 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
143 ACL. Previously, a crash would result.
145 JH/19 Speed up macro lookups during configuration file read, by skipping non-
146 macro text after a replacement (previously it was only once per line) and
147 by skipping builtin macros when searching for an uppercase lead character.
149 JH/20 DANE support moved from Experimental to mainline. The Makefile control
150 for the build is renamed.
152 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
153 was allocated for every new TLS startup, meaning one per message. Fix
154 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
156 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
157 reported the original. Fix to report (as far as possible) the ACL
158 result replacing the original.
160 JH/23 Fix memory leak during multi-message connections using STARTTLS under
161 OpenSSL. Certificate information is loaded for every new TLS startup,
162 and the resources needed to be freed.
164 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
166 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
169 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
170 DATA response info to the (existing) per-recipient response info for
171 the "C=" log element. It can have useful tracking info from the
172 destination system. Patch from Simon Arlott.
174 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
175 length value. Previously this would segfault.
177 HS/02 Support Avast multiline protoocol, this allows passing flags to
178 newer versions of the scanner.
180 JH/28 Ensure that variables possibly set during message acceptance are marked
181 dead before release of memory in the daemon loop. This stops complaints
182 about them when the debug_store option is enabled. Discovered specifically
183 for sender_rate_period, but applies to a whole set of variables.
184 Do the same for the queue-runner and queue-list loops, for variables set
185 from spool message files. Do the same for the SMTP per-message loop, for
186 certain variables indirectly set in ACL operations.
188 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
189 as a multi-recipient message from a mailinglist manager). The coding had
190 an arbitrary cutoff number of characters while checking for more input;
191 enforced by writing a NUL into the buffer. This corrupted long / fast
192 input. The problem was exposed more widely when more pipelineing of SMTP
193 responses was introduced, and one Exim system was feeding another.
194 The symptom is log complaints of SMTP syntax error (NUL chars) on the
195 receiving system, and refused recipients seen by the sending system
196 (propating to people being dropped from mailing lists).
197 Discovered and pinpointed by David Carter.
199 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
200 replaced by the ${authresults } expansion.
202 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
204 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
205 allows proper process termination in container environments.
207 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
208 Previously the "final dot" had a newline after it; ensure it is CR,LF.
210 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
211 and "err_perm", deprecated since 4.83 when the RFC-defined words
212 "temperror" and "permerror" were introduced.
214 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
215 transport-filters or DKIM-signing. The restriction was lost in the
216 consolidation of verify-callout and delivery SMTP handling.
217 Extend the restriction to also cover ARC-signing.
219 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
220 in defer=pass mode supply a 450 to the initiator. Previously the message
223 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
224 tls_require_ciphers is used as before.
226 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
227 "pass_unscanned". Only tmpfails from the scanner are written to
228 the paniclog, as they may require admin intervention (permission
229 denied, license issues). Other scanner errors (like decompression
230 bombs) do not cause a paniclog entry.
232 JH/36 Fix reinitialisation of DKIM logging variable between messages.
233 Previously it was possible to log spurious information in receive log
236 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
237 triggered odd behaviour from Outlook Express clients.
239 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
242 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
243 since the IETF WG has not yet settled on that versus the original
244 "bare" representation.
246 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
247 Previously the millisecond value corrupted the output.
248 Fix also for syslog_pid=no and log_selector +pid, for which the pid
249 corrupted the output.
255 JH/01 Rework error string handling in TLS interface so that the caller in
256 more cases is responsible for logging. This permits library-sourced
257 string to be attached to addresses during delivery, and collapses
258 pairs of long lines into single ones.
260 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
261 during configuration. Wildcards are allowed and expanded.
263 JH/02 Rework error string handling in DKIM to pass more info back to callers.
264 This permits better logging.
266 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
267 do not close it down and have the child transport start it up again on
268 the passed-on TCP connection. Instead, proxy the child (and any
269 subsequent ones) for TLS via a unix-domain socket channel. Logging is
270 affected: the continued delivery log lines do not have any DNSSEC, TLS
271 Certificate or OCSP information. TLS cipher information is still logged.
273 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
274 identical IP addresses on different listening ports. Will also affect
277 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
278 add noisy ifdef guards to special-case this sillyness.
279 Patch from Bernd Kuhls.
281 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
282 than 255 are no longer allowed.
284 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
285 Disable the session-cache too, which might reduce our load. Since we
286 currrectly use a new context for every connection, both as server and
287 client, there is no benefit for these.
288 GnuTLS appears to not support tickets server-side by default (we don't
289 call gnutls_session_ticket_enable_server()) but client side is enabled
290 by default on recent versions (3.1.3 +) unless the PFS priority string
293 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
294 <https://reproducible-builds.org/specs/source-date-epoch/>.
296 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
297 the check for any unsuccessful recipients did not notice the limit, and
298 erroneously found still-pending ones.
300 JH/08 Pipeline CHUNKING command and data together, on kernels that support
301 MSG_MORE. Only in-clear (not on TLS connections).
303 JH/09 Avoid using a temporary file during transport using dkim. Unless a
304 transport-filter is involved we can buffer the headers in memory for
305 creating the signature, and read the spool data file once for the
306 signature and again for transmission.
308 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
309 4.77 as the kernel support then wasn't solid, having issues in 64bit
310 mode. Now, it's been long enough. Add support for FreeBSD also.
312 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
313 case where the routing stage had gathered several addresses to send to
314 a host before calling the transport for the first, we previously failed
315 to close down TLS in the old transport process before passing the TCP
316 connection to the new process. The new one sent a STARTTLS command
317 which naturally failed, giving a failed delivery and bloating the retry
318 database. Investigation and fix prototype from Wolfgang Breyha.
320 JH/12 Fix check on SMTP command input synchronisation. Previously there were
321 false-negatives in the check that the sender had not preempted a response
322 or prompt from Exim (running as a server), due to that code's lack of
323 awareness of the SMTP input buffering.
325 PP/04 Add commandline_checks_require_admin option.
326 Exim drops privileges sanely, various checks such as -be aren't a
327 security problem, as long as you trust local users with access to their
328 own account. When invoked by services which pass untrusted data to
329 Exim, this might be an issue. Set this option in main configuration
330 AND make fixes to the calling application, such as using `--` to stop
333 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
334 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
335 responses to those, into a single TLS record each way (this usually means
336 a single packet). As a side issue, smtp_enforce_sync now works on TLS
339 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
340 affects you only if you're dancing at the edge of the param size limits.
341 If you are, and this message makes sense to you, then: raise the
342 configured limit or use OpenSSL 1.1. Nothing we can do for older
345 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
346 cmdline element to get the documented default one. Previously it was
349 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
350 are unsuable for non-message-specific reasons - eg. connection timeout,
351 banner-time rejection.
353 JH/16 Fix logging of delivery remote port, when specified by router, under
356 PP/06 Repair manualroute's ability to take options in any order, even if one
357 is the name of a transport.
360 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
362 JH/17 Change the list-building routines interface to use the expanding-string
363 triplet model, for better allocation and copying behaviour.
365 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
366 Previously it was constructed the first time a possibly-matching string
367 was met in the configuration file input during startup; now it is done
370 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
371 compatible one, to avoid the (poorly documented) possibility of a config
372 file in the working directory redirecting the DB files, possibly correpting
373 some existing file. CVE-2017-10140 assigned for BDB.
375 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
376 cache-hot. Previously, although the result was properly cached, the
377 initial verify call returned a defer.
379 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
380 the main verify for receipient in uncached-mode.
382 JH/22 Retire historical build files to an "unsupported" subdir. These are
383 defined as "ones for which we have no current evidence of testing".
385 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
386 if present. Previously it was ignored.
388 JH/24 Start using specified-initialisers in C structure init coding. This is
389 a C99 feature (it's 2017, so now considered safe).
391 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
392 if was a fixed-sized field and bitmask ops via macros; it is now more
395 PP/07 GitHub PR 56: Apply MariaDB build fix.
396 Patch provided by Jaroslav Škarvada.
398 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
399 during Coverity cleanups [4.87 JH/47]
400 Diagnosis and fix provided by Michael Fischer v. Mollard.
402 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
403 the right size to place the terminating semicolon on its own folded
404 line, the header hash was calculated to an incorrect value thanks to
405 the (relaxed) space the fold became.
407 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
408 and confused the parent.
410 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
411 which could crash as a result. This could lead to undeliverable messages.
413 JH/28 Logging: "next input sent too soon" now shows where input was truncated
416 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
417 matters on fast-turnover and PID-randomising systems, which were getting
418 out-of-order delivery.
420 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
421 a possibly-overlapping copy. The symptom was that "Remote host closed
422 connection in response to HELO" was logged instead of the actual 4xx
425 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
426 Previously only that bufferd was discarded, resulting in SYMTP command
429 JH/32 DKIM: when a message has multiple signatures matching an identity given
430 in dkim_verify_signers, run the dkim acl once for each. Previously only
431 one run was done. Bug 2189.
433 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
434 "panic the current process" to "deliberately defer". The panic log is
435 still written with the problem list name; the mail and reject logs now
436 get a temp-reject line for the message that was being handled, saying
437 something like "domains check lookup or other defer". The SMTP 451
438 message is still "Temporary local problem".
440 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
441 A crafted sequence of BDAT commands could result in in-use memory beeing
442 freed. CVE-2017-16943.
444 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
445 from SMTP input. Previously it was always done; now only done for DATA
446 and not BDAT commands. CVE-2017-16944.
448 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
449 to the message (such as an overlong header line). Previously this was
450 not done and we did not exit BDAT mode. Followon from the previous item
451 though a different problem.
457 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
458 than -2003 did; needs libidn2 in addition to libidn.
460 JH/02 The path option on a pipe transport is now expanded before use.
462 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
463 Patch provided by "Björn", documentation fix added too.
465 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
466 missing a wire-to-host endian conversion.
468 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
469 close after a BDAT command line could be taken as a following command,
470 giving a synch failure. Fix by only checking for synch immediately
471 before acknowledging the chunk.
473 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
474 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
475 macro. Patches provided by Josh Soref.
477 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
478 Previously we did not; the RFC seems ambiguous and VRFY is not listed
479 by IANA as a service extension. However, John Klensin suggests that we
482 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
483 the dkim code may be unix-mode line endings rather than smtp wire-format
484 CRLF, so prepend a CR to any bare LF.
486 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
487 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
489 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
490 the first were themselves being wrongly included in the feed into dkim
491 processing; with most chunk sizes in use this resulted in an incorrect
492 body hash calculated value.
494 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
495 DKIM signature block, for verification. Although advised against by
496 standards it is specifically not ruled illegal.
498 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
500 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
501 missing a body hash (the bh= tag).
503 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
504 It seems that HAProxy sends the Proxy Protocol information in clear and
505 only then does a TLS startup, so do the same.
507 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
508 TCP connections (such as for Spamd) unless the daemon successfully set
509 Fast Open mode on its listening sockets. This fixes breakage seen on
510 too-old kernels or those not configured for Fast Open, at the cost of
511 requiring both directions being enabled for TFO, and TFO never being used
512 by non-daemon-related Exim processes.
514 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
515 endings, at least on the first header line. Try to canonify any that get
516 past that check, despite the cost.
518 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
519 now limited to an arbitrary five deep, while parsing addresses with the
520 strip_excess_angle_brackets option enabled.
522 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
523 instead leave the unprompted TLS handshake in socket buffer for the
524 TLS library to consume.
526 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
528 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
530 JH/16 Drop variables when they go out of scope. Memory management drops a whole
531 region in one operation, for speed, and this leaves assigned pointers
532 dangling. Add checks run only under the testsuite which checks all
533 variables at a store-reset and panics on a dangling pointer; add code
534 explicitly nulling out all the variables discovered. Fixes one known
535 bug: a transport crash, where a dangling pointer for $sending_ip_address
536 originally assigned in a verify callout, is re-used.
538 PP/06 Drop '.' from @INC in various Perl scripts.
540 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
542 PP/08 Reduce a number of compilation warnings under clang; building with
543 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
544 should be warning-free.
546 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
548 HS/01 Fix portability problems introduced by PP/08 for platforms where
549 realloc(NULL) is not equivalent to malloc() [SunOS et al].
551 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
552 chunk. This allows us to accept broken chunked messages. We need a more
553 general solution here.
555 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
556 already-broken messages in the queue.
558 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
560 JH/19 Fix reference counting bug in routing-generated-address tracking.
566 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
567 supports it and a size is available (ie. the sending peer gave us one).
569 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
570 years of being deprecated). The replacements are the ACLs
571 acl_smtp_mime and acl_not_smtp_mime.
573 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
574 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
575 or even an in-clear connection were permitted. Now, if the host lookup
576 was dnssec and dane was requested then the host is only used if the
577 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
578 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
579 if one fails this test.
580 This means that a poorly-configured remote DNS will make it incommunicado;
581 but it protects against a DNS-interception attack on it.
583 JH/04 Bug 1810: make continued-use of an open smtp transport connection
584 non-noisy when a race steals the message being considered.
586 JH/05 If main configuration option tls_certificate is unset, generate a
587 self-signed certificate for inbound TLS connections.
589 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
590 in rewrites and routers.
592 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
593 and logged a warning sing 4.83; now they are a configuration file error.
595 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
596 (lacking @domain). Apply the same qualification processing as RCPT.
598 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
600 JH/10 Support ${sha256:} applied to a string (as well as the previous
603 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
604 a cutthrough deliver is pending, as we always want to make a connection.
605 This also avoids re-routing the message when later placing the cutthrough
606 connection after a verify cache hit.
607 Do not update it with the verify result either.
609 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
610 when routing results in more than one destination address.
612 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
613 signing (which inhibits the cutthrough capability). Previously only
614 the presence of an option was tested; now an expansion evaluating as
615 empty is permissible (obviously it should depend only on data available
616 when the cutthrough connection is made).
618 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
619 the relevant preceding SMTP command did not note the pipelining mode.
621 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
622 Previously they were not counted.
624 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
625 as one having no matching records. Previously we deferred the message
626 that needed the lookup.
628 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
629 distinguished as "(=".
631 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
632 for missing MX records. Previously it only worked for missing A records.
634 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
636 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
637 after the data-go-ahead and data-ack. Patch from Jason Betts.
639 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
640 even for a "none" policy. Patch from Tony Meyer.
642 JH/22 Fix continued use of a connection for further deliveries. If a port was
643 specified by a router, it must also match for the delivery to be
646 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
647 When one of the recipients of a message was unsuitable for the connection
648 (has no matching addresses), we lost track of needing to mark it
649 deferred. As a result mail would be lost.
651 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
653 JH/25 Decoding ACL controls is now done using a binary search; the source code
654 takes up less space and should be simpler to maintain. Merge the ACL
655 condition decode tables also, with similar effect.
657 JH/26 Fix problem with one_time used on a redirect router which returned the
658 parent address unchanged. A retry would see the parent address marked as
659 delivered, so not attempt the (identical) child. As a result mail would
662 JH/27 Fix a possible security hole, wherein a process operating with the Exim
663 UID can gain a root shell. Credit to http://www.halfdog.net/ for
664 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
667 JH/28 Enable {spool,log} filesystem space and inode checks as default.
668 Main config options check_{log,spool}_{inodes,space} are now
669 100 inodes, 10MB unless set otherwise in the configuration.
671 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
672 Previously it only applied to the main-section connection policy
675 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
677 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
678 by me. Added RFC7919 DH primes as an alternative.
680 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
681 are not in the system include path.
683 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
684 GnuTLS, when a session startup failed (eg because the client disconnected)
685 Exim did stdio operations after fclose. This was exposed by a recent
686 change which nulled out the file handle after the fclose.
688 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
689 signed directly by the cert-signing cert, rather than an intermediate
690 OCSP-signing cert. This is the model used by LetsEncrypt.
692 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
694 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
695 an incoming connection.
697 HS/02 Bug 1802: Do not half-close the connection after sending a request
700 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
701 fallback to "prime256v1".
703 JH/34 SECURITY: Use proper copy of DATA command in error message.
704 Could leak key material. Remotely exploitable. CVE-2016-9963.
710 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
711 and 3.4.4 - once the server is enabled to respond to an OCSP request
712 it does even when not requested, resulting in a stapling non-aware
713 client dropping the TLS connection.
715 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
716 support variable-length bit vectors. No functional change.
718 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
719 The I= interface field on outgoing lines is now after the H= remote
720 host field, same as incoming lines. There is a separate
721 outgoing_interface log selector which allows you to disable the
724 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
725 If not running log_selector +smtp_connection the mainlog would be held
726 open indefinitely after a "too many connections" event, including to a
727 deleted file after a log rotate. Leave the per net connection logging
728 leaving it open for efficiency as that will be quickly detected by the
729 check on the next write.
731 HS/01 Bug 1671: Fix post transport crash.
732 Processing the wait-<transport> messages could crash the delivery
733 process if the message IDs didn't exist for some reason. When
734 using 'split_spool_directory=yes' the construction of the spool
735 file name failed already, exposing the same netto behaviour.
737 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
738 mime_regex ACL conditions.
740 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
741 to DSN fail messages (bounces): remote IP, remote greeting, remote response
742 to HELO, local diagnostic string.
744 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
745 log line to debug. Even when configured with a tls authenticator many
746 client connections are expected to not authenticate in this way, so
747 an authenticate fail is not an error.
749 HS/02 Add the Exim version string to the process info. This way exiwhat
750 gives some more detail about the running daemon.
752 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
753 matter for fast-change records such as DNSBLs.
755 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
756 retry record, even if constant. There may be multiple transports with
757 different interface settings and the retry behaviour needs to be kept
760 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
762 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
764 JH/10 Bug 840: fix log_defer_output option of pipe transport
766 JH/11 Bug 830: use same host for all RCPTS of a message, even under
767 hosts_randomize. This matters a lot when combined with mua_wrapper.
769 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
770 ${quote_pgsql:<string>} operator.
772 JH/13 Bug 1708: avoid misaligned access in cached lookup.
774 JH/14 Change header file name for freeradius-client. Relevant if compiling
775 with Radius support; from the Gentoo tree and checked under Fedora.
777 JH/15 Bug 1712: Introduce $prdr_requested flag variable
779 JH/16 Bug 1714: Permit an empty string as expansion result for transport
780 option transport_filter, meaning no filtering.
782 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
784 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
785 defaults to "*" (all hosts). The variable is now available when not built
786 with TLS, default unset, mainly to enable keeping the testsuite sane.
787 If a server certificate is not supplied (via tls_certificate) an error is
788 logged, and clients will find TLS connections fail on startup. Presumably
789 they will retry in-clear.
790 Packagers of Exim are strongly encouraged to create a server certificate
791 at installation time.
793 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
794 with the $config_file variable.
796 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
797 in transport context, after the attempt, and per-recipient. The latter type
798 is per host attempted. The event data is the error message, and the errno
799 information encodes the lookup type (A vs. MX) used for the (first) host,
800 and the trailing two digits of the smtp 4xx response.
802 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
803 to write to mainlog (or rejectlog, paniclog) in the window between file
804 creation and permissions/ownership being changed. Particularly affects
805 installations where exicyclog is run as root, rather than exim user;
806 result is that the running daemon panics and dies.
808 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
810 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
811 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
812 "pri" and "weight". Note that the previous implicit priority given by the
813 list order is no longer honoured.
815 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
818 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
819 by defining SUPPORT_SOCKS.
821 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
822 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
823 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
824 variables are renamed to proxy_{local,external}_{address,port}.
826 JH/27 Move Internationalisation support from Experimental to mainline, enabled
827 for a build by defining SUPPORT_I18N
829 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
830 of the query string, and make ${quote_redis:} do that quoting.
832 JH/29 Move Events support from Experimental to mainline, enabled by default
833 and removable for a build by defining DISABLE_EVENT.
835 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
837 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
838 cached by the daemon.
840 JH/32 Move Redis support from Experimental to mainline, enabled for a build
841 by defining LOOKUP_REDIS. The libhiredis library is required.
843 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
844 keys are given for lookup.
846 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
847 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
848 only supported when built with TLS support. The PolarSSL SHA routines
849 are still used when the TLS library is too old for convenient support.
851 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
852 openssl_options), for security. OpenSSL forces this from version 1.1.0
853 server-side so match that on older versions.
855 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
856 allocation for $value could be released as the expansion processing
857 concluded, but leaving the global pointer active for it.
859 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
860 and to use the domains and local_parts ACL conditions.
862 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
863 incorrectly not doubled on cutthrough transmission, hence seen as a
864 body-termination at the receiving system - resulting in truncated mails.
865 Commonly the sender saw a TCP-level error, and retransmitted the message
866 via the normal store-and-forward channel. This could result in duplicates
867 received - but deduplicating mailstores were liable to retain only the
868 initial truncated version.
870 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
872 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
874 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
875 we're in there, support oversigning also; bug 1309.
877 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
879 HS/04 Add support for keep_environment and add_environment options.
881 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
882 either intentional arithmetic overflow during PRNG, or testing config-
885 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
886 delivery resulted in actual delivery. Cancel cutthrough before DATA
889 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
890 rejects a recipient: pass the reject to the originator.
892 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
893 Many were false-positives and ignorable, but it's worth fixing the
896 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
897 for the new environment-manipulation done at startup. Move the routines
898 from being local to tls.c to being global via the os.c file.
900 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
901 an extract embedded as result-arg for a map, the first arg for extract
902 is unavailable so we cannot tell if this is a numbered or keyed
903 extraction. Accept either.
909 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
912 JH/02 The smtp transport option "multi_domain" is now expanded.
914 JH/03 The smtp transport now requests PRDR by default, if the server offers
917 JH/04 Certificate name checking on server certificates, when exim is a client,
918 is now done by default. The transport option tls_verify_cert_hostnames
919 can be used to disable this per-host. The build option
920 EXPERIMENTAL_CERTNAMES is withdrawn.
922 JH/05 The value of the tls_verify_certificates smtp transport and main options
923 default to the word "system" to access the system default CA bundle.
924 For GnuTLS, only version 3.0.20 or later.
926 JH/06 Verification of the server certificate for a TLS connection is now tried
927 (but not required) by default. The verification status is now logged by
928 default, for both outbound TLS and client-certificate supplying inbound
931 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
934 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
935 Status Notification (bounce) messages are now MIME format per RFC 3464.
936 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
937 under the control of the dsn_advertise_hosts option, and routers may
938 have a dsn_lasthop option.
940 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
941 default, modifiable by a malware= option. The list separator for
942 the options can now be changed in the usual way. Bug 68.
944 JH/10 The smtp_receive_timeout main option is now expanded before use.
946 JH/11 The incoming_interface log option now also enables logging of the
947 local interface on delivery outgoing connections.
949 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
950 if the interface and destination host and port all match.
952 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
955 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
956 Patch from Andrew Lewis.
958 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
959 now supports optional time-restrictions, weighting, and priority
960 modifiers per server. Patch originally by <rommer@active.by>.
962 JH/16 The spamd_address main option now supports a mixed list of local
963 and remote servers. Remote servers can be IPv6 addresses, and
964 specify a port-range.
966 JH/17 Bug 68: The spamd_address main option now supports an optional
967 timeout value per server.
969 JH/18 Bug 1581: Router and transport options headers_add/remove can
970 now have the list separator specified.
972 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
975 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
978 JH/21 Support for the A6 type of dns record is withdrawn.
980 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
981 rather than the verbs used.
983 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
984 from 255 to 1024 chars.
986 JH/24 Verification callouts now attempt to use TLS by default.
988 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
989 are generic router options now. The defaults didn't change.
991 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
992 Original patch from Alexander Shikoff, worked over by JH.
994 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
995 files if ZCAT_COMMAND is not executable.
997 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
999 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1001 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1002 Normally benign, it bites when the pair was led to by a CNAME;
1003 modern usage is to not canonicalize the domain to a CNAME target
1004 (and we were inconsistent anyway for A-only vs AAAA+A).
1006 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1008 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1009 when evaluating $sender_host_dnssec.
1011 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1012 $sender_helo_dnssec variable.
1014 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1016 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1018 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1020 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1021 documented as working, but never had. Support all but $spam_report.
1023 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1024 added for tls authenticator.
1026 HS/03 Add perl_taintmode main config option
1032 TL/01 When running the test suite, the README says that variables such as
1033 no_msglog_check are global and can be placed anywhere in a specific
1034 test's script, however it was observed that placement needed to be near
1035 the beginning for it to behave that way. Changed the runtest perl
1036 script to read through the entire script once to detect and set these
1037 variables, reset to the beginning of the script, and then run through
1038 the script parsing/test process like normal.
1040 TL/02 The BSD's have an arc4random API. One of the functions to induce
1041 adding randomness was arc4random_stir(), but it has been removed in
1042 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1043 function when detected.
1045 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1046 cause callback expansion.
1048 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1049 syntax errors in an expansion can be treated as a string instead of
1050 logging or causing an error, due to the internal use of bool_lax
1051 instead of bool when processing it.
1053 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1054 server certificates when making smtp deliveries.
1056 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1058 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1060 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1062 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1063 Merged patch from Sebastian Wiedenroth.
1065 JH/05 Fix results-pipe from transport process. Several recipients, combined
1066 with certificate use, exposed issues where response data items split
1067 over buffer boundaries were not parsed properly. This eventually
1068 resulted in duplicates being sent. This issue only became common enough
1069 to notice due to the introduction of connection certificate information,
1070 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1072 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1073 size buffer was used, resulting in syntax errors when an expansion
1076 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1077 version 3.3.6 or later.
1079 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1080 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1081 both become "event_action", the variables become $event_name, $event_data
1082 and $event_defer_errno. There is a new variable $verify_mode, usable in
1083 routers, transports and related events. The tls:cert event is now also
1084 raised for inbound connections, if the main configuration event_action
1087 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1088 early OCSP support, but no stapling (appears to be less than 1.0.0).
1090 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1091 server certificate names available under the smtp transport option
1092 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1095 JH/10 Time-related extraction expansions from certificates now use the main
1096 option "timezone" setting for output formatting, and are consistent
1097 between OpenSSL and GnuTLS compilations. Bug 1541.
1099 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1100 encoded parameter in the incoming message. Bug 1558.
1102 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1103 include certificate info, eximon was claiming there were spoolfile
1106 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1108 JH/14 Log delivery-related information more consistently, using the sequence
1109 "H=<name> [<ip>]" wherever possible.
1111 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1112 are problematic for Debian distribution, omit them from the release
1115 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1117 JH/16 Fix string representation of time values on 64bit time_t architectures.
1120 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1121 output list separator was used.
1126 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1127 checkers that were complaining about end of non-void function with no
1130 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1131 This was a regression introduced in 4.83 by another bugfix.
1133 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1135 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1136 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1142 TF/01 Correctly close the server side of TLS when forking for delivery.
1144 When a message was received over SMTP with TLS, Exim failed to clear up
1145 the incoming connection properly after forking off the child process to
1146 deliver the message. In some situations the subsequent outgoing
1147 delivery connection happened to have the same fd number as the incoming
1148 connection previously had. Exim would try to use TLS and fail, logging
1149 a "Bad file descriptor" error.
1151 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1152 utilities have not been installed.
1154 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1155 temporary space as the ACL may create new global variables.
1157 TL/01 LDAP support uses per connection or global context settings, depending
1158 upon the detected version of the libraries at build time.
1160 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1161 to extract and use the src ip:port in logging and expansions as if it
1162 were a direct connection from the outside internet. PPv2 support was
1163 updated based on HAProxy spec change in May 2014.
1165 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1167 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1168 Properly escape header and check for NULL return.
1170 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1173 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1175 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1176 characters in header names, implemented as a verify condition.
1177 Contributed by Michael Fischer v. Mollard.
1179 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1180 results permerror and temperror. Previous values are deprecated but
1181 still accepted. In a future release, err_perm and err_temp will be
1182 completely removed, which will be a backward incompatibility if the
1183 ACL tests for either of these two old results. Patch contributed by
1184 user bes-internal on the mailing list.
1186 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1188 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1189 selectors, in both main and reject logs.
1191 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1194 JH/07 Add malware type "sock" for talking to simple daemon.
1196 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1198 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1199 routers/transports under cutthrough routing.
1201 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1202 numbers. Touch up "bool" conditional to keep the same definition.
1204 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1206 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1208 JH/12 Expand items in router/transport headers_add or headers_remove lists
1209 individually rather than the list as a whole. Bug 1452.
1211 Required for reasonable handling of multiple headers_ options when
1212 they may be empty; requires that headers_remove items with embedded
1213 colons must have them doubled (or the list-separator changed).
1215 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1216 view the policy declared in the DMARC record. Currently, $dmarc_status
1217 is a combined value of both the record presence and the result of the
1220 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1222 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1223 dnslookup router and the smtp transport (applying to the forward
1226 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1227 of ldap servers used for a specific lookup. Patch provided by Heiko
1230 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1231 New variable $lookup_dnssec_authenticated for observability.
1233 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1234 Patch submitted by Lars Timman.
1236 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1238 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1239 Requires trusted mode and valid format message id, aborts otherwise.
1240 Patch contributed by Heiko Schlichting.
1242 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1243 certextract with support for various fields. Bug 1358.
1245 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1246 is requested by default, modifiable by smtp transport option
1249 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1250 operate on certificate variables to give certificate fingerprints
1251 Also new ${sha256:cert_variable}.
1253 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1255 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1256 Christian Aistleitner.
1258 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1260 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1261 file. Patch from Wolfgang Breyha.
1263 JH/25 Expand the coverage of the delivery $host and $host_address to
1264 client authenticators run in verify callout. Bug 1476.
1266 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1267 align with daemon_smtp_ports. Bug 72.
1269 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1270 support and error reporting did not work properly.
1272 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1273 and is readable. Patch from Andrew Colin Kissa.
1275 TL/14 Enhance documentation of ${run expansion and how it parses the
1276 commandline after expansion, particularly in the case when an
1277 unquoted variable expansion results in an empty value.
1279 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1281 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1282 by renaming away from it. Addresses GH issue 15, reported by
1285 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1287 TL/15 SECURITY: prevent double expansion in math comparison functions
1288 (can expand unsanitized data). Not remotely exploitable.
1295 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1297 PP/02 Make -n do something, by making it not do something.
1298 When combined with -bP, the name of an option is not output.
1300 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1303 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1304 $sender_host_name and config options to manage this, and basic check
1307 PP/05 DSCP support for outbound connections and control modifier for inbound.
1309 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1310 (Only plugin which currently uses this is kerberos4, which nobody should
1311 be using, but we should make it available and other future plugins might
1312 conceivably use it, even though it would break NAT; stuff *should* be
1313 using channel bindings instead).
1315 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1316 name; added for Sendmail compatibility; requires admin caller.
1317 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1318 just ignore it); requires trusted caller.
1319 Also parse but ignore: -Ac -Am -X<logfile>
1322 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1324 TL/02 Add +smtp_confirmation as a default logging option.
1326 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1327 Patch by Magnus Holmgren from 2007-02-20.
1329 TL/04 Bugzilla 1281 - Spec typo.
1330 Bugzilla 1283 - Spec typo.
1331 Bugzilla 1290 - Spec grammar fixes.
1333 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1335 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1337 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1338 mailing list by by Dmitry Isaikin.
1340 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1342 JH/02 Support "G" suffix to numbers in ${if comparisons.
1344 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1346 NM/01 Bugzilla 1197 - Spec typo
1347 Bugzilla 1196 - Spec examples corrections
1349 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1351 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1352 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1355 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1356 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1358 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1359 "acl {{name}{arg}...}", and optional args on acl condition
1362 JH/05 Permit multiple router/transport headers_add/remove lines.
1364 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1366 JH/07 Avoid using a waiting database for a single-message-only transport.
1367 Performance patch from Paul Fisher. Bugzilla 1262.
1369 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1372 JH/09 Add $headers_added variable, with content from use of ACL modifier
1373 add_header (but not yet added to the message). Bugzilla 199.
1375 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1376 Pulled from Bugzilla 817 by Wolfgang Breyha.
1378 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1380 (nb: this is the same fix as in Exim 4.80.1)
1382 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1385 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1388 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1390 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1391 advertise SMTP AUTH mechanism to us, instead of a generic
1392 protocol violation error. Also, make Exim more robust to bad
1393 data from the Dovecot auth socket.
1395 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1397 When a queue runner is handling a message, Exim first routes the
1398 recipient addresses, during which it prunes them based on the retry
1399 hints database. After that it attempts to deliver the message to
1400 any remaining recipients. It then updates the hints database using
1403 So if a recipient address works intermittently, it can get repeatedly
1404 deferred at routing time. The retry hints record remains fresh so the
1405 address never reaches the final cutoff time.
1407 This is a fairly common occurrence when a user is bumping up against
1408 their storage quota. Exim had some logic in its local delivery code
1409 to deal with this. However it did not apply to per-recipient defers
1410 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1412 This change adds a proper retry rule check during routing so that the
1413 final cutoff time is checked against the message's age. We only do
1414 this check if there is an address retry record and there is not a
1415 domain retry record; this implies that previous attempts to handle
1416 the address had the retry_use_local_parts option turned on. We use
1417 this as an approximation for the destination being like a local
1418 delivery, as in LMTP.
1420 I suspect this new check makes the old local delivery cutoff check
1421 redundant, but I have not verified this so I left the code in place.
1423 TF/02 Correct gecos expansion when From: is a prefix of the username.
1425 Test 0254 submits a message to Exim with the header
1429 When I ran the test suite under the user fanf2, Exim expanded
1430 the header to contain my full name, whereas it should have added
1431 a Resent-Sender: header. It erroneously treats any prefix of the
1432 username as equal to the username.
1434 This change corrects that bug.
1436 GF/01 DCC debug and logging tidyup
1437 Error conditions log to paniclog rather than rejectlog.
1438 Debug lines prefixed by "DCC: " to remove any ambiguity.
1440 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1442 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1443 Bug spotted by Jeremy Harris; was flawed since initial commit.
1444 Would have resulted in OCSP responses post-SNI triggering an Exim
1445 NULL dereference and crash.
1447 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1449 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1450 Bug detection, analysis and fix by Samuel Thibault.
1451 Bugzilla 1331, Debian bug #698092.
1453 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1455 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1456 Server implementation by Todd Lyons, client by JH.
1457 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1458 config variable "prdr_enable" controls whether the server
1459 advertises the facility. If the client requests PRDR a new
1460 acl_data_smtp_prdr ACL is called once for each recipient, after
1461 the body content is received and before the acl_smtp_data ACL.
1462 The client is controlled by both of: a hosts_try_prdr option
1463 on the smtp transport, and the server advertisement.
1464 Default client logging of deliveries and rejections involving
1465 PRDR are flagged with the string "PRDR".
1467 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1468 fclose(). Diagnosis by Todd Lyons.
1470 PP/17 Update configure.default to handle IPv6 localhost better.
1471 Patch by Alain Williams (plus minor tweaks).
1474 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1475 This is now consistent with GnuTLS, and is now documented: the
1476 previous undocumented portable approach to treating the option as
1477 unset was to force an expansion failure. That still works, and
1478 an empty string is now equivalent.
1480 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1481 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1482 not performing validation itself.
1484 PP/20 Added force_command boolean option to pipe transport.
1485 Patch from Nick Koston, of cPanel Inc.
1487 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1490 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1492 PP/21 Fix eximon continuous updating with timestamped log-files.
1493 Broken in a format-string cleanup in 4.80, missed when I repaired the
1494 other false fix of the same issue.
1495 Report and fix from Heiko Schlichting.
1498 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1499 Report from Prashanth Katuri.
1501 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1502 It's SecureTransport, so affects any MacOS clients which use the
1503 system-integrated TLS libraries, including email clients.
1505 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1506 using a MIME ACL for non-SMTP local injection.
1507 Report and assistance in diagnosis by Warren Baker.
1509 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1511 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1513 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1514 last id that failed so it may be referenced in subsequent ACL's.
1516 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1519 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1520 ldap library initialization, allowing self-signed CA's to be
1521 used. Also properly sets require_cert option later in code by
1522 using NULL (global ldap config) instead of ldap handle (per
1523 session). Bug diagnosis and testing by alxgomz.
1525 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1526 the src/util/ subdirectory.
1528 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1529 renamed to Transport Post Delivery Action by Jeremy Harris, as
1532 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1533 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1534 redis_servers = needs to be configured which will be used by the redis
1535 lookup. Patch from Warren Baker, of The Packet Hub.
1537 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1539 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1540 hostname or reverse DNS when processing a host list. Used suggestions
1541 from multiple comments on this bug.
1543 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1545 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1546 Missed a few lines, added it to make the runtest require no keyboard
1549 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1550 contains upper case chars. Make router use caseful_local_part.
1552 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1553 support when GnuTLS has been built with p11-kit.
1559 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1561 This, or similar/improved, will also be change PP/11 of 4.82.
1567 PP/01 Handle short writes when writing local log-files.
1568 In practice, only affects FreeBSD (8 onwards).
1569 Bugzilla 1053, with thanks to Dmitry Isaikin.
1571 NM/01 Bugzilla 949 - Documentation tweak
1573 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1576 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1578 PP/02 Implemented gsasl authenticator.
1580 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1582 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1583 `pkg-config foo` for cflags/libs.
1585 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1586 with rest of GSASL and with heimdal_gssapi.
1588 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1589 `pkg-config foo` for cflags/libs for the TLS implementation.
1591 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1592 properties get this fed in as external SSF. A number of robustness
1593 and debugging improvements to the cyrus_sasl authenticator.
1595 PP/08 cyrus_sasl server now expands the server_realm option.
1597 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1598 Patch by Jeremy Harris.
1600 PP/10 Added dbmjz lookup type.
1602 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1604 PP/12 MAIL args handles TAB as well as SP, for better interop with
1605 non-compliant senders.
1606 Analysis and variant patch by Todd Lyons.
1608 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1609 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1610 Patch from Dirk Mueller <dmueller@suse.com>
1612 PP/13 tls_peerdn now print-escaped for spool files.
1613 Observed some $tls_peerdn in wild which contained \n, which resulted
1614 in spool file corruption.
1616 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1617 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1618 or write after TLS renegotiation, which otherwise led to messages
1621 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1622 as a tracking header (ie: a signed header comes before the signature).
1623 Patch from Wolfgang Breyha.
1625 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1626 comma-sep list; embedded commas doubled.
1628 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1630 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1632 Report and patch from Dmitry Banschikov.
1634 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1635 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1636 failures. We appear to now support TLS1.1+ with Exim.
1638 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1639 lets Exim select keys and certificates based upon TLS SNI from client.
1640 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1641 before an outbound SMTP session. New log_selector, +tls_sni.
1643 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1644 NULL dereference. Report and patch from Alun Jones.
1646 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1647 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1648 Not seeing resolver debug output on NetBSD, but suspect this is a
1649 resolver implementation change.
1651 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1652 Left warnings. Added "eximon gdb" invocation mode.
1654 PP/21 Defaulting "accept_8bitmime" to true, not false.
1656 PP/22 Added -bw for inetd wait mode support.
1658 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1659 locate the relevant includes and libraries. Made this the default.
1661 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1662 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1664 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1665 This may cause build issues on older platforms.
1667 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1668 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1669 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1670 Added SNI support via GnuTLS too.
1671 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1673 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1675 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1676 Applied second patch from Janne, implementing suggestion to default
1677 multiple-strings-in-record handling to match SPF spec.
1679 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1681 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1682 read-only, out of scope).
1683 Patch from Wolfgang Breyha, report from Stuart Northfield.
1685 PP/29 Fix three issues highlighted by clang analyser static analysis.
1686 Only crash-plausible issue would require the Cambridge-specific
1687 iplookup router and a misconfiguration.
1688 Report from Marcin Mirosław.
1690 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1692 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1693 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1694 As part of this, removing so much warning spew let me fix some minor
1695 real issues in debug logging.
1697 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1698 assignment on my part. Fixed.
1700 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1701 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1702 Janne Snabb (who went above and beyond: thank you).
1704 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1705 string otherwise requires a connection and a bunch more work and it's
1706 relatively easy to get wrong. Should also expose TLS library linkage
1709 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1710 64-bit ${eval} (JH/03).
1712 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1713 GNU libc to support some of the 64-bit stuff, should not lead to
1714 conflicts. Defined before os.h is pulled in, so if a given platform
1715 needs to override this, it can.
1717 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1718 protection layer was required, which is not implemented.
1719 Bugzilla 1254, patch from Wolfgang Breyha.
1721 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1722 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1723 tls_dhparam take prime identifiers. Also unbreak combination of
1724 OpenSSL+DH_params+TLSSNI.
1726 PP/39 Disable SSLv2 by default in OpenSSL support.
1732 PP/01 Solaris build fix for Oracle's LDAP libraries.
1733 Bugzilla 1109, patch from Stephen Usher.
1735 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1737 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1740 TF/02 Fix a couple more cases where we did not log the error message
1741 when unlink() failed. See also change 4.74-TF/03.
1743 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1744 lock up or crash if it happened to be inside a call to libc when it
1745 got a SIGUSR1 from exiwhat.
1747 The SIGUSR1 handler appends the current process status to the process
1748 log which is later printed by exiwhat. It used to use the general
1749 purpose logging code to do this, but several functions it calls are
1750 not safe for signals.
1752 The new output code in the SIGUSR1 handler is specific to the process
1753 log, and simple enough that it's easy to inspect for signal safety.
1754 Removing some special cases also simplifies the general logging code.
1755 Removing the spurious timestamps from the process log simplifies
1758 TF/04 Improved ratelimit ACL condition.
1760 The /noupdate option has been deprecated in favour of /readonly which
1761 has clearer semantics. The /leaky, /strict, and /readonly update modes
1762 are mutually exclusive. The update mode is no longer included in the
1763 database key; it just determines when the database is updated. (This
1764 means that when you upgrade Exim will forget old rate measurements.)
1766 Exim now checks that the per_* options are used with an update mode that
1767 makes sense for the current ACL. For example, when Exim is processing a
1768 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1769 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1770 must specify per_mail/readonly. If you omit the update mode it defaults to
1771 /leaky where that makes sense (as before) or /readonly where required.
1773 The /noupdate option is now undocumented but still supported for
1774 backwards compatibility. It is equivalent to /readonly except that in
1775 ACLs where /readonly is required you may specify /leaky/noupdate or
1776 /strict/noupdate which are treated the same as /readonly.
1778 A useful new feature is the /count= option. This is a generalization
1779 of the per_byte option, so that you can measure the throughput of other
1780 aggregate values. For example, the per_byte option is now equivalent
1781 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1783 The per_rcpt option has been generalized using the /count= mechanism
1784 (though it's more complicated than the per_byte equivalence). When it is
1785 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1786 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1787 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1788 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1789 Note that using per_rcpt with a non-readonly update mode in more than
1790 one ACL will cause the recipients to be double-counted. (The per_mail
1791 and per_byte options don't have this problem.)
1793 The handling of very low rates has changed slightly. If the computed rate
1794 is less than the event's count (usually one) then this event is the first
1795 after a long gap. In this case the rate is set to the same as this event's
1796 count, so that the first message of a spam run is counted properly.
1798 The major new feature is a mechanism for counting the rate of unique
1799 events. The new per_addr option counts the number of different
1800 recipients that someone has sent messages to in the last time period. It
1801 behaves like per_rcpt if all the recipient addresses are different, but
1802 duplicate recipient addresses do not increase the measured rate. Like
1803 the /count= option this is a general mechanism, so the per_addr option
1804 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1805 example, measure the rate that a client uses different sender addresses
1806 with the options per_mail/unique=$sender_address. There are further
1807 details in the main documentation.
1809 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1811 TF/06 Removed a few PCRE remnants.
1813 TF/07 Automatically extract Exim's version number from tags in the git
1814 repository when doing development or release builds.
1816 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1817 Bugzilla 879. Patch from Paul Fisher.
1819 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1820 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1823 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1825 PP/05 Variable $av_failed, true if the AV scanner deferred.
1826 Bugzilla 1078. Patch from John Horne.
1828 PP/06 Stop make process more reliably on build failure.
1829 Bugzilla 1087. Patch from Heiko Schlittermann.
1831 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1832 Bugzilla 1089. Patch from Heiko Schlittermann.
1834 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1835 Bugzilla 1131. Patch from Holger Weiß.
1837 PP/09 Handle IPv6 addresses with SPF.
1838 Bugzilla 860. Patch from Wolfgang Breyha.
1840 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1842 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1845 PP/11 match_* no longer expand right-hand-side by default.
1846 New compile-time build option, EXPAND_LISTMATCH_RHS.
1847 New expansion conditions, "inlist", "inlisti".
1849 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1851 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1853 PP/14 fix log_write() format string regression from TF/03.
1854 Bugzilla 1152. Patch from Dmitry Isaikin.
1860 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1862 PP/02 Harmonised TLS library version reporting; only show if debugging.
1863 Layout now matches that introduced for other libraries in 4.74 PP/03.
1865 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1867 PP/04 New "dns_use_edns0" global option.
1869 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1872 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1873 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1875 TK/01 Updated PolarSSL code to 0.14.2.
1876 Bugzilla 1097. Patch from Andreas Metzler.
1878 PP/07 Catch divide-by-zero in ${eval:...}.
1879 Fixes bugzilla 1102.
1881 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1884 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1885 format-string attack -- SECURITY: remote arbitrary code execution.
1887 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1888 time unintentionally subject to list matching rules, letting the header
1889 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1890 arbitrary string expansion). This allowed for information disclosure.
1892 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1893 INT_MIN/-1 -- value coerced to INT_MAX.
1899 NM/01 Workaround for PCRE version dependency in version reporting
1902 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1903 This fixes portability to compilers other than gcc, notably
1904 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1906 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1907 makefiles for portability to HP-UX and POSIX correctness.
1909 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1910 Also via indented variable definition in the Makefile.
1911 (Debugging by Oliver Heesakkers).
1913 PP/02 Restore caching of spamd results with expanded spamd_address.
1914 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1916 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1917 Improves build reliability. Fix from: Frank Elsner
1919 NM/02 Fix wide character breakage in the rfc2047 coding
1920 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1922 NM/03 Allow underscore in dnslist lookups
1923 Fixes bug 1026. Patch from Graeme Fowler
1925 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1926 Code patches from Adam Ciarcinski of NetBSD.
1928 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1931 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1932 is logged, to avoid truncation. Patch from John Horne.
1934 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1935 Patch from Jakob Hirsch.
1937 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1938 SQL string expansion failure details.
1939 Patch from Andrey Oktyabrski.
1941 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1942 Patch from Simon Arlott.
1944 PP/09 New lookups functionality failed to compile on old gcc which rejects
1945 extern declarations in function scope.
1946 Patch from Oliver Fleischmann
1948 PP/10 Use sig_atomic_t for flags set from signal handlers.
1949 Check getgroups() return and improve debugging.
1950 Fixed developed for diagnosis in bug 927 (which turned out to be
1953 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1954 Patch from Mark Zealey.
1956 PP/12 Bugzilla 1056: Improved spamd server selection.
1957 Patch from Mark Zealey.
1959 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1960 Based on patch from Heiko Schlittermann.
1962 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1963 Patch from Uwe Doering, sign-off by Michael Haardt.
1965 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1966 variable declaration deep within a block. Bug and patch from
1969 PP/15 lookups-Makefile IRIX compatibility coercion.
1971 PP/16 Make DISABLE_DKIM build knob functional.
1973 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1974 Patch by Simon Arlott
1976 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1977 variable argument macros. Our copy now differs from upstream.
1983 TF/01 Failure to get a lock on a hints database can have serious
1984 consequences so log it to the panic log.
1986 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1987 controlled using the smtp_confirmation log selector.
1989 TF/03 Include the error message when we fail to unlink a spool file.
1991 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1992 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1993 for maintaining out-of-tree patches for some time.
1995 PP/01 Bugzilla 139: Documentation and portability issues.
1996 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1997 Handle per-OS dynamic-module compilation flags.
1999 PP/02 Let /dev/null have normal permissions.
2000 The 4.73 fixes were a little too stringent and complained about the
2001 permissions on /dev/null. Exempt it from some checks.
2002 Reported by Andreas M. Kirchwitz.
2004 PP/03 Report version information for many libraries, including
2005 Exim version information for dynamically loaded libraries. Created
2006 version.h, now support a version extension string for distributors
2007 who patch heavily. Dynamic module ABI change.
2009 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2010 privilege escalation vulnerability whereby the Exim run-time user
2011 can cause root to append content of the attacker's choosing to
2014 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2017 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2018 If dropping privileges for untrusted macros, we disabled normal logging
2019 on the basis that it would fail; for the Exim run-time user, this is not
2020 the case, and it resulted in successful deliveries going unlogged.
2021 Fixed. Reported by Andreas Metzler.
2027 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2028 only prepend for the Resent-* case. Fixes regression introduced in
2029 Exim 4.70 by NM/22 for Bugzilla 607.
2031 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2032 increasing numbers of administrators be bitten by this.
2034 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2036 PP/03 Bugzilla 994: added openssl_options main configuration option.
2038 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2040 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2042 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2044 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2045 without a peer certificate, leading to a segfault because of an
2046 assumption that peers always have certificates. Be a little more
2047 paranoid. Problem reported by Martin Tscholak.
2049 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2050 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2051 NB: ClamAV planning to remove STREAM in "middle of 2010".
2052 CL also introduces -bmalware, various -d+acl logging additions and
2053 more caution in buffer sizes.
2055 PP/09 Implemented reverse_ip expansion operator.
2057 PP/10 Bugzilla 937: provide a "debug" ACL control.
2059 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2061 PP/12 Bugzilla 973: Implement --version.
2063 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2065 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2067 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2069 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2070 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2071 ignore trailing whitespace.
2073 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2075 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2078 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2079 Notification from Dr Andrew Aitchison.
2081 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2082 ExtendedDetectionInfo response format.
2083 Notification from John Horne.
2085 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2088 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2089 XSL and documented dependency on system catalogs, with examples of how
2092 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2095 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2096 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2097 configuration file which is writeable by the Exim user or group.
2099 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2100 of configuration files to cover files specified with the -C option if
2101 they are going to be used with root privileges, not just the default
2104 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2105 option (effectively making it always true).
2107 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2108 files to be used while preserving root privileges.
2110 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2111 that rogue child processes cannot use them.
2113 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2114 run-time user, instead of root.
2116 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2117 Exim run-time user without dropping privileges.
2119 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2120 result string, instead of calling string_vformat() twice with the same
2123 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2124 for other users. Others should always drop root privileges if they use
2125 -C on the command line, even for a whitelisted configure file.
2127 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2129 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2135 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2136 $data_path, and $header_path variables; fixed documentation bugs and
2139 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2140 exipick to access non-standard spools, including the "frozen" queue
2143 NM/01 Bugzilla 965: Support mysql stored procedures.
2144 Patch from Alain Williams
2146 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2148 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2149 Patch from Andreas Metzler
2151 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2152 Patch from Kirill Miazine
2154 NM/05 Bugzilla 671: Added umask to procmail example.
2156 JJ/03 installed exipick 20100323.0, fixing doc bug
2158 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2159 directory. Notification and patch from Dan Rosenberg.
2161 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2163 TK/02 Improve log output when DKIM signing operation fails.
2165 MH/01 Treat the transport option dkim_domain as a colon separated
2166 list, not as a single string, and sign the message with each element,
2167 omitting multiple occurences of the same signer.
2169 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2170 Bugzilla 985, 986. Patch by Simon Arlott
2172 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2173 Patch by Simon Arlott
2175 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2176 MBX locking. Notification from Dan Rosenberg.
2182 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2184 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2186 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2188 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2190 NM/04 Bugzilla 915: Flag broken perl installation during build.
2196 TK/01 Added patch by Johannes Berg that expands the main option
2197 "spamd_address" if it starts with a dollar sign.
2199 TK/02 Write list of recipients to X-Envelope-Sender header when building
2200 the mbox-format spool file for content scanning (suggested by Jakob
2203 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2204 (http://www.dcc-servers.net/) support via dccifd. Activated by
2205 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2207 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2208 by Mark Daniel Reidel <mr@df.eu>.
2210 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2211 When building exim an external PCRE library is now needed -
2212 PCRE is a system library on the majority of modern systems.
2213 See entry on PCRE_LIBS in EDITME file.
2215 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2216 conversation. Added nologin parameter to request.
2217 Patch contributed by Kirill Miazine.
2219 TF/01 Do not log submission mode rewrites if they do not change the address.
2221 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2223 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2224 log files in place. Contributed by Roberto Lima.
2226 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2228 TF/03 Bugzilla 615: When checking the local_parts router precondition
2229 after a local_part_suffix or local_part_prefix option, Exim now
2230 does not use the address's named list lookup cache, since this
2231 contains cached lookups for the whole local part.
2233 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2234 Robert Millan. Documentation is in experimental-spec.txt.
2236 TF/04 Bugzilla 668: Fix parallel build (make -j).
2238 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2240 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2241 Patch provided by Jan Srzednicki.
2243 TF/05 Leading white space used to be stripped from $spam_report which
2244 wrecked the formatting. Now it is preserved.
2246 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2247 that they are available at delivery time.
2249 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2251 TF/08 TLS error reporting now respects the incoming_interface and
2252 incoming_port log selectors.
2254 TF/09 Produce a more useful error message if an SMTP transport's hosts
2255 setting expands to an empty string.
2257 NM/06 Bugzilla 744: EXPN did not work under TLS.
2258 Patch provided by Phil Pennock.
2260 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2261 Patch provided by Richard Godbee.
2263 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2264 acl_smtp_notquit, added index entry.
2266 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2267 Patch provided by Eugene Bujak.
2269 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2270 accept(). Patch provided by Maxim Dounin.
2272 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2273 Patch provided by Phil Pennock.
2275 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2277 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2278 Patch provided by Brad "anomie" Jorsch.
2280 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2281 Patch provided by Dean Brooks.
2283 TK/05 Add native DKIM support (does not depend on external libraries).
2285 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2286 Patch provided by Graeme Fowler.
2288 NM/16 Bugzilla 851: Documentation example syntax fix.
2290 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2292 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2295 NM/19 Bugzilla 745: TLS version reporting.
2296 Patch provided by Phil Pennock.
2298 NM/20 Bugzilla 167: bool: condition support.
2299 Patch provided by Phil Pennock.
2301 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2302 clients. Patch provided by Phil Pennock.
2304 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2305 Patch provided by Brad "anomie" Jorsch.
2307 NM/23 Bugzilla 687: Fix misparses in eximstats.
2308 Patch provided by Heiko Schlittermann.
2310 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2311 Patch provided by Heiko Schlittermann.
2313 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2314 plus update to original patch.
2316 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2318 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2319 Patch provided by David Brownlee.
2321 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2323 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2325 NM/30 Bugzilla 888: TLS documentation bugfixes.
2327 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2329 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2330 Unlike the original bugzilla I have changed all shell scripts in src tree.
2332 NM/33 Bugzilla 898: Transport filter timeout fix.
2333 Patch by Todd Rinaldo.
2335 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2336 Patch by Serge Demonchaux.
2338 NM/35 Bugzilla 39: Base64 decode bug fixes.
2339 Patch by Jakob Hirsch.
2341 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2343 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2345 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2347 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2353 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2354 ALT-N's libdkim that I have put here:
2355 http://duncanthrax.net/exim-experimental/
2357 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2358 were called 'true' and it seems that C99 defines that as a reserved
2359 keyword to be used with 'bool' variable types. That means you could
2360 not include C99-style headers which use bools without triggering
2361 build errors in sieve.c.
2363 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2364 as mailq or other aliases. Changed the --help handling significantly
2365 to do whats expected. exim_usage() emits usage/help information.
2367 SC/01 Added the -bylocaldomain option to eximstats.
2369 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2371 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2373 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2379 PH/01 Another patch from the Sieve maintainer.
2381 PH/02 When an IPv6 address is converted to a string for single-key lookup
2382 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2383 dots are used instead of colons so that keys in lsearch files need not
2384 contain colons. This was done some time before quoting was made available
2385 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2386 (notated using the quote facility) so as to distinguish them from IPv4
2387 keys. This meant that lookups for IP addresses in host lists did not work
2388 for iplsearch lookups.
2390 This has been fixed by arranging for IPv6 addresses to be expressed with
2391 colons if the lookup type is iplsearch. This is not incompatible, because
2392 previously such lookups could never work.
2394 The situation is now rather anomalous, since one *can* have colons in
2395 ordinary lsearch keys. However, making the change in all cases is
2396 incompatible and would probably break a number of configurations.
2398 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2401 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2402 conversion specification without a maximum field width, thereby enabling
2403 a rogue spamd server to cause a buffer overflow. While nobody in their
2404 right mind would setup Exim to query an untrusted spamd server, an
2405 attacker that gains access to a server running spamd could potentially
2406 exploit this vulnerability to run arbitrary code as the Exim user.
2408 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2409 $primary_hostname instead of what libspf2 thinks the hosts name is.
2411 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2412 a directory entry by the name of the lookup key. Previously, if a
2413 symlink pointed to a non-existing file or a file in a directory that
2414 Exim lacked permissions to read, a lookup for a key matching that
2415 symlink would fail. Now it is enough that a matching directory entry
2416 exists, symlink or not. (Bugzilla 503.)
2418 PH/03 The body_linecount and body_zerocount variables are now exported in the
2421 PH/04 Added the $dnslist_matched variable.
2423 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2424 This means they are set thereafter only if the connection becomes
2427 PH/06 Added the client_condition to authenticators so that some can be skipped
2428 by clients under certain conditions.
2430 PH/07 The error message for a badly-placed control=no_multiline_responses left
2431 "_responses" off the end of the name.
2433 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2435 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2436 (without spaces) instead of just copying the configuration text.
2438 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2440 PH/11 Added $max_received_linelength.
2442 PH/12 Added +ignore_defer and +include_defer to host lists.
2444 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2445 way in which PCRE > 7.0 is built.
2447 PH/14 Implemented queue_only_load_latch.
2449 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2450 MAIL command. The effect was to mangle the value on 64-bit systems.
2452 PH/16 Another patch from the Sieve maintainer.
2454 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2456 PH/18 If a system quota error occurred while trying to create the file for
2457 a maildir delivery, the message "Mailbox is full" was not appended to the
2458 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2459 only to a quota excession during the actual writing of the file.
2461 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2462 characters?) which causes problems in log lines. The DN values are now
2463 passed through string_printing() before being added to log lines.
2465 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2466 and InterBase are left for another time.)
2468 PH/21 Added message_body_newlines option.
2470 PH/22 Guard against possible overflow in moan_check_errorcopy().
2472 PH/23 POSIX allows open() to be a macro; guard against that.
2474 PH/24 If the recipient of an error message contained an @ in the local part
2475 (suitably quoted, of course), incorrect values were put in $domain and
2476 $local_part during the evaluation of errors_copy.
2482 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2483 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2486 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2487 issue a MAIL command.
2489 PH/02 In an ACL statement such as
2491 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2493 if a client was not listed at all, or was listed with a value other than
2494 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2495 the condition was not true (as it should be), so access was not denied.
2496 The bug was that the ! inversion was incorrectly passed on to the second
2497 item. This has been fixed.
2499 PH/03 Added additional dnslists conditions == and =& which are different from
2500 = and & when the dns lookup returns more than one IP address.
2502 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2503 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2505 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2506 FSYNC, which compiles an option called disable_fsync that allows for
2507 bypassing fsync(). The documentation is heavily laced with warnings.
2509 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2511 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2512 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2513 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2514 including adding "make clean"; (3) Added -fPIC when compiling the test
2515 dynamically loaded module, to get rid of a warning.
2517 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2518 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2519 = 0s. The bug is otherwise harmless.
2521 PH/07 There was a bug in the dovecot authenticator such that the value of
2522 $auth1 could be overwritten, and so not correctly preserved, after a
2523 successful authentication. This usually meant that the value preserved by
2524 the server_setid option was incorrect.
2526 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2528 PH/09 Installed PCRE release 7.0.
2530 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2531 run for batched SMTP input. It is now run at the start of every message
2532 in the batch. While fixing this I discovered that the process information
2533 (output by running exiwhat) was not always getting set for -bs and -bS
2534 input. This is fixed, and it now also says "batched" for BSMTP.
2536 PH/11 Added control=no_pipelining.
2538 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2539 patch, slightly modified), and move the expansion of helo_data till after
2540 the connection is made in the smtp transport (so it can use these
2543 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2545 PH/14 Added log_selector = +pid.
2547 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2549 PH/16 Add ${if forany and ${if forall.
2551 PH/17 Added dsn_from option to vary the From: line in DSNs.
2553 PH/18 Flush SMTP output before performing a callout, unless control =
2554 no_callout_flush is set.
2556 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2557 was true (the default) a successful delivery failed to delete the retry
2558 item, thus causing premature timeout of the address. The bug is now
2561 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2563 PH/21 Long custom messages for fakedefer and fakereject are now split up
2564 into multiline responses in the same way that messages for "deny" and
2565 other ACL rejections are.
2567 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2568 with slight modification.
2570 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2571 draft, changing the syntax and factoring some duplicate code".
2573 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2574 for deliveries of the second and subsequent messages over the same SMTP
2577 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2578 ${reduce, with only minor "tidies".
2580 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2582 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2583 expansion side effects.
2585 PH/27 When a message times out after an over-quota error from an Exim-imposed
2586 quota, the bounce message says "mailbox is full". This message was not
2587 being given when it was a system quota that was exceeded. It now should
2590 MH/03 Made $recipients available in local_scan(). local_scan() already has
2591 better access to the recipient list through recipients_list[], but
2592 $recipients can be useful in postmaster-provided expansion strings.
2594 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2595 in the case of a MAIL command with additional options following the
2596 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2597 were accidentally chopped off.
2599 PH/29 SMTP synchronization checks are implemented when a command is read -
2600 there is a check that no more input is waiting when there shouldn't be
2601 any. However, for some commands, a delay in an ACL can mean that it is
2602 some time before the response is written. In this time, more input might
2603 arrive, invalidly. So now there are extra checks after an ACL has run for
2604 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2605 pipelining has not been advertised.
2607 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2609 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2610 returned in the SMTP response when smtp_return_error_details was false.
2611 This has been fixed.
2613 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2614 instead of the C I/O that was originally supplied, because problems were
2615 reported on Solaris.
2617 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2618 Exim which did not show up earlier: it was assuming that a call to
2619 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2620 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2621 was a macro that became an assignment, so it seemed to work. This has
2622 changed to a proper function call with a void return, hence the compile
2623 error. Exim's code has been fixed.
2625 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2628 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2630 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2632 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2633 because it turns out to be too big or there is a timeout) there is no
2634 "Completed" line in the log. When some messages of this type were
2635 selected by exigrep, they were listed as "not completed". Others were
2636 picked up by some special patterns. I have improved the selection
2637 criteria to be more general.
2639 PH/38 The host_find_failed option in the manualroute router can now be set
2640 to "ignore", to completely ignore a host whose IP address cannot be
2641 found. If all hosts are ignored, the behaviour is controlled by the new
2642 host_all_ignored option.
2644 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2645 homing or because of multiple MX records with /mx) generated more than
2646 one IP address, and the following item turned out to be the local host,
2647 all the secondary addresses of the first item were incorrectly removed
2648 from the list, along with the local host and any following hosts (which
2649 is what is supposed to happen).
2651 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2652 whoever called Exim into the -H file. In the case of the daemon it was
2653 behaving confusingly. When first started, it used values for whoever
2654 started the daemon, but after a SIGHUP it used the Exim user (because it
2655 calls itself on a restart). I have changed the code so that it now always
2658 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2659 message are rejected with the same error (e.g. no authentication or bad
2660 sender address), and a DATA command is nevertheless sent (as can happen
2661 with PIPELINING or a stupid MUA), the error message that was given to the
2662 RCPT commands is included in the rejection of the DATA command. This is
2663 intended to be helpful for MUAs that show only the final error to their
2666 PH/42 Another patch from the Sieve maintainer.
2668 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2669 Eximstats - Fixed some broken HTML links and added missing column headers
2671 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2672 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2674 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2676 PH/43 Yet another patch from the Sieve maintainer.
2678 PH/44 I found a way to check for a TCP/IP connection going away before sending
2679 the response to the final '.' that terminates a message, but only in the
2680 case where the client has not sent further data following the '.'
2681 (unfortunately, this is allowed). However, in many cases there won't be
2682 any further data because there won't be any more messages to send. A call
2683 to select() can be used: if it shows that the input is "ready", there is
2684 either input waiting, or the socket has been closed. An attempt to read
2685 the next input character can distinguish the two cases. Previously, Exim
2686 would have sent an OK response which the client would never have see.
2687 This could lead to message repetition. This fix should cure that, at
2688 least in a lot of common cases.
2690 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2691 advertised in response to EHLO.
2697 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2698 fixed by 4.65/MH/01 (is this a record?) are fixed:
2700 (i) An empty string was always treated as zero by the numeric comparison
2701 operators. This behaviour has been restored.
2703 (ii) It is documented that the numeric comparison operators always treat
2704 their arguments as decimal numbers. This was broken in that numbers
2705 starting with 0 were being interpreted as octal.
2707 While fixing these problems I realized that there was another issue that
2708 hadn't been noticed. Values of message_size_limit (both the global option
2709 and the transport option) were treated as octal if they started with 0.
2710 The documentation was vague. These values are now always treated as
2711 decimal, and I will make that clear in the documentation.
2717 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2718 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2721 MH/01 Don't check that the operands of numeric comparison operators are
2722 integers when their expansion is in "skipping" mode (fixes bug
2723 introduced by 4.64-PH/07).
2725 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2726 child addresses, Exim now panics and dies. Previously, because the count
2727 is held in a short int, deliveries were likely to be lost. As such a
2728 large number of recipients for a single message is ridiculous
2729 (performance will be very, very poor), I have chosen to impose a limit
2730 rather than extend the field.
2736 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2737 leftover -K file (the existence of which was triggered by #402).
2738 While we were at it, introduced process PID as part of the -K
2739 filename. This should rule out race conditions when creating
2742 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2743 processing considerably. Previous code took too long for large mails,
2744 triggering a timeout which in turn triggers #401.
2746 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2747 in the DK code in transports.c. sendfile() is not really portable,
2748 hence the _LINUX specificness.
2750 TF/01 In the add_headers option to the mail command in an Exim filter,
2751 there was a bug that Exim would claim a syntax error in any
2752 header after the first one which had an odd number of characters
2755 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2756 callout verification, Exim cached a "reject" for the entire domain. This
2757 is correct for most verifications, but it is not correct for a recipient
2758 verification with use_sender or use_postmaster set, because in that case
2759 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2760 case of MAIL FROM:<> rejection from other early rejections (e.g.
2761 rejection of HELO). When verifying a recipient using a non-null MAIL
2762 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2763 Whatever the result of the callout, the value of the domain cache is
2764 left unchanged (for any other kind of callout, getting as far as trying
2765 RCPT means that the domain itself is ok).
2767 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2770 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2771 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2772 session, the daemon ignores ECONNECT errors and logs others; it now
2773 ignores EPIPE as well.
2775 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2776 (quoted-printable decoding).
2778 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2779 later the small subsequent patch to fix an introduced bug.
2781 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2783 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2785 PH/08 An error is now given if message_size_limit is specified negative.
2787 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2788 to be given (somewhat) arbitrary names.
2790 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2793 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2794 miscellaneous code fixes
2796 PH/10 Added the log_reject_target ACL modifier to specify where to log
2799 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2800 hostname. This is wrong, because it relates to the incoming message (and
2801 probably the interface on which it is arriving) and not to the outgoing
2802 callout (which could be using a different interface). This has been
2803 changed to use the value of the helo_data option from the smtp transport
2804 instead - this is what is used when a message is actually being sent. If
2805 there is no remote transport (possible with a router that sets up host
2806 addresses), $smtp_active_hostname is used.
2808 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2809 tweaks were necessary in order to get it to work (see also 21 below):
2810 (a) The code assumed that strncpy() returns a negative number on buffer
2811 overflow, which isn't the case. Replaced with Exim's string_format()
2813 (b) There were several signed/unsigned issues. I just did the minimum
2814 hacking in of casts. There is scope for a larger refactoring.
2815 (c) The code used strcasecmp() which is not a standard C function.
2816 Replaced with Exim's strcmpic() function.
2817 (d) The code set only $1; it now sets $auth1 as well.
2818 (e) A simple test gave the error "authentication client didn't specify
2819 service in request". It would seem that Dovecot has changed its
2820 interface. Fortunately there's a specification; I followed it and
2821 changed what the client sends and it appears to be working now.
2823 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2826 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2827 address A is aliased to B and C, where B exists and C does not. Without
2828 -v the output is "A verified" because verification stops after a
2829 successful redirection if more than one address is generated. However,
2830 with -v the child addresses are also verified. Exim was outputting "A
2831 failed to verify" and then showing the successful verification for C,
2832 with its parentage. It now outputs "B failed to verify", showing B's
2833 parentage before showing the successful verification of C.
2835 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2836 look up a TXT record in a specific list after matching in a combined
2839 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2840 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2841 they consult the DNS. I had assumed they would set it the way they
2842 wanted; and indeed my experiments on Linux seem to show that in some
2843 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2844 To be on the safe side, however, I have now made the interface to
2845 host_find_byname() similar to host_find_bydns(), with an argument
2846 containing the DNS resolver options. The host_find_byname() function now
2847 sets these options at its start, just as host_find_bydns() does. The smtp
2848 transport options dns_qualify_single and dns_search_parents are passed to
2849 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2850 of host_find_byname() use the default settings of RES_DEFNAMES
2851 (qualify_single) but not RES_DNSRCH (search_parents).
2853 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2854 spool_read_header() do less string testing, by means of a preliminary
2855 switch on the second character of optional "-foo" lines. (This is
2856 overdue, caused by the large number of possibilities that now exist.
2857 Originally there were few.) While I was there, I also converted the
2858 str(n)cmp tests so they don't re-test the leading "-" and the first
2859 character, in the hope this might squeeze out yet more improvement.
2861 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2862 flag allowing group syntax was set by the header_syntax check but not
2863 turned off, possible causing trouble later; (2) The flag was not being
2864 set at all for the header_verify test, causing "group"-style headers to
2865 be rejected. I have now set it in this case, and also caused header_
2866 verify to ignore an empty address taken from a group. While doing this, I
2867 came across some other cases where the code for allowing group syntax
2868 while scanning a header line wasn't quite right (mostly, not resetting
2869 the flag correctly in the right place). These bugs could have caused
2870 trouble for malformed header lines. I hope it is now all correct.
2872 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2873 with the "reply" argument non-NULL. The code, however (which originally
2874 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2875 but it didn't always do it. This confused somebody who was copying the
2876 code for some other use. I have removed all the tests.
2878 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2879 feature that was used to support insecure browsers during the U.S. crypto
2880 embargo. It requires special client support, and Exim is probably the
2881 only MTA that supported it -- and would never use it because real RSA is
2882 always available. This code has been removed, because it had the bad
2883 effect of slowing Exim down by computing (never used) parameters for the
2884 RSA_EXPORT functionality.
2886 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2887 authenticator to fail if there's a tab character in the incoming data
2888 (there should never be unless someone is messing about, as it's supposed
2889 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2890 if the connection is using TLS or if the remote IP is the same as the
2891 local IP, and the "valid-client-cert option" if a client certificate has
2894 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2895 authenticators. This can be used for authorization after authentication
2896 succeeds. (In the case of plaintext, it servers for both authentication
2899 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2900 if any retry times were supplied.
2902 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2903 connection, where there is no client IP address to check. In this
2904 situation, the verify now always succeeds.
2906 PH/25 Applied John Jetmore's -Mset patch.
2908 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2910 PH/27 In a string expansion for a processed (not raw) header when multiple
2911 headers of the same name were present, leading whitespace was being
2912 removed from all of them, but trailing whitespace was being removed only
2913 from the last one. Now trailing whitespace is removed from each header
2914 before concatenation. Completely empty headers in a concatenation (as
2915 before) are ignored.
2917 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2918 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2920 PH/29 [Removed. This was a change that I later backed out, and forgot to
2921 correct the ChangeLog entry (that I had efficiently created) before
2922 committing the later change.]
2924 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2925 address errors (4xx response to RCPT) over the same connection as other
2926 messages routed to the same hosts. Such deliveries are always "forced",
2927 so retry times are not inspected. This resulted in far too many retries
2928 for the affected addresses. The effect occurred only when there were more
2929 hosts than the hosts_max_try setting in the smtp transport when it had
2930 the 4xx errors. Those hosts that it had tried were not added to the list
2931 of hosts for which the message was waiting, so if all were tried, there
2932 was no problem. Two fixes have been applied:
2934 (i) If there are any address or message errors in an SMTP delivery, none
2935 of the hosts (tried or untried) are now added to the list of hosts
2936 for which the message is waiting, so the message should not be a
2937 candidate for sending over the same connection that was used for a
2938 successful delivery of some other message. This seems entirely
2939 reasonable: after all the message is NOT "waiting for some host".
2940 This is so "obvious" that I'm not sure why it wasn't done
2941 previously. Hope I haven't missed anything, but it can't do any
2942 harm, as the worst effect is to miss an optimization.
2944 (ii) If, despite (i), such a delivery is accidentally attempted, the
2945 routing retry time is respected, so at least it doesn't keep
2946 hammering the server.
2948 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2949 in ${readsocket because some servers need this prod.
2951 PH/32 Added some extra debug output when updating a wait-xxx database.
2953 PH/33 The hint "could be header name not terminated by colon", which has been
2954 given for certain expansion errors for a long time, was not being given
2955 for the ${if def:h_colon_omitted{... case.
2957 PH/34 The spec says: "With one important exception, whenever a domain list is
2958 being scanned, $domain contains the subject domain." There was at least
2959 one case where this was not true.
2961 PH/35 The error "getsockname() failed: connection reset by peer" was being
2962 written to the panic log as well as the main log, but it isn't really
2963 panic-worthy as it just means the connection died rather early on. I have
2964 removed the panic log writing for the ECONNRESET error when getsockname()
2967 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2968 runs only) independently of the message's sender address. This meant
2969 that, if the 4xx error was in fact related to the sender, a different
2970 message to the same recipient with a different sender could confuse
2971 things. In particular, this can happen when sending to a greylisting
2972 server, but other circumstances could also provoke similar problems.
2973 I have changed the default so that the retry time for these errors is now
2974 based a combination of the sender and recipient addresses. This change
2975 can be overridden by setting address_retry_include_sender=false in the
2978 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2979 remote server are returned as part of bounce messages. This was not
2980 happening for LMTP over a pipe (the lmtp transport), but now it is the
2981 same for both kinds of LMTP.
2983 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2984 in header lines that were in fact CNAMEs. This is no longer the case.
2986 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2987 and queue runs started by the daemon processed all messages. This has
2988 been fixed so that -R and -S can now usefully be given with -q<time>.
2990 PH/40 Import PCRE release 6.7 (fixes some bugs).
2992 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2994 PH/42 Give an error if -q is specified more than once.
2996 PH/43 Renamed the variables $interface_address and $interface_port as
2997 $received_ip_address and $received_port, to make it clear that these
2998 values apply to message reception, and not to the outgoing interface when
2999 a message is delivered. (The old names remain recognized, of course.)
3001 PH/44 There was no timeout on the connect() call when using a Unix domain
3002 socket in the ${readsocket expansion. There now is.
3004 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3005 be meaningful with "accept".
3007 SC/01 Eximstats V1.43
3008 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3010 SC/02 Eximstats V1.44
3011 Use a glob alias rather than an array ref in the generated
3012 parser. This improves both readability and performance.
3014 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3015 Collect SpamAssassin and rejection statistics.
3016 Don't display local sender or destination tables unless
3017 there is data to show.
3018 Added average volumes into the top table text output.
3020 SC/04 Eximstats V1.46
3021 Collect data on the number of addresses (recipients)
3022 as well as the number of messages.
3024 SC/05 Eximstats V1.47
3025 Added 'Message too big' to the list of mail rejection
3026 reasons (thanks to Marco Gaiarin).
3028 SC/06 Eximstats V1.48
3029 Mainlog lines which have GMT offsets and are too short to
3030 have a flag are now skipped.
3032 SC/07 Eximstats V1.49 (Alain Williams)
3033 Added the -emptyok flag.
3035 SC/08 Eximstats V1.50
3036 Fixes for obtaining the IP address from reject messages.
3038 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3039 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3040 whitespace changes from 4.64-PH/27
3042 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3045 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3046 are found, allow negative numbers in numeric criteria)
3048 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3050 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3051 to match changes made in 4.64-PH/43
3053 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3055 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3056 "domain" is actually an IP address, and force a failure. This locks out
3057 those revolvers/nameservers that support "A-for-A" lookups, in
3058 contravention of the specifications.
3060 PH/48 When a host name was looked up from an IP address, and the subsequent
3061 forward lookup of the name timed out, the host name was left in
3062 $sender_host_name, contrary to the specification.
3064 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3065 restricted to single-key lookups, Exim was not diagnosing an error if
3066 * or *@ was used with a query-style lookup.
3068 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3070 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3071 long ago, but noone interested enough thought of it. Let's just say that
3072 the "1.1" means that there are some new functions that weren't there at
3073 some point in the past.
3075 PH/51 Error processing for expansion failure of helo_data from an smtp
3076 transport during callout processing was broken.
3078 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3079 tested/used via the -bh/-bhc/-bs options.
3081 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3082 bug, fixed in subsequent PCRE releases).
3084 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3085 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3087 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3093 SC/01 Use a glob alias rather than an array ref in eximstats generated
3094 parser. This improves both readability and performance.
3096 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3097 Don't display local sender or destination tables in eximstats unless
3098 there is data to show.
3099 Added average volumes into the eximstats top table text output.
3101 SC/03 Collect data on the number of addresses (recipients) as well
3102 as the number of messages in eximstats.
3104 TF/01 Correct an error in the documentation for the redirect router. Exim
3105 does (usually) call initgroups() when daemonizing.
3107 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3108 with consistent privilege compared to when running as a daemon.
3110 TF/03 Note in the spec that $authenticated_id is not set for local
3111 submissions from trusted users.
3113 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3114 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3116 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3117 by adding some example configuration directives to the default
3118 configuration file. A little bit of work is required to uncomment the
3119 directives and define how usernames and passwords are checked, but
3120 there is now a framework to start from.
3122 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3123 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3124 without this. I don't know how relevant this is to other LDAP libraries.
3126 PH/02 Add the verb name to the "unknown ACL verb" error.
3128 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3130 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3132 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3133 directory not expanded when it should be if an expanded home directory
3134 was set for the address (which is overridden by the transport).
3136 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3139 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3140 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3141 because it is too late at that time, and has no effect.
3143 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3144 security issue with \' (bugzilla #107). I could not use the
3145 PQescapeStringConn() function, because it needs a PGconn value as one of
3148 PH/08 When testing addresses using -bt, indicate those final addresses that
3149 are duplicates that would not cause an additional delivery. At least one
3150 person was confused, thinking that -bt output corresponded to deliveries.
3151 (Suppressing duplicates isn't a good idea as you lose the information
3152 about possibly different redirections that led to the duplicates.)
3154 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3155 systems where poll() doesn't work, in particular OS X.
3157 PH/10 Added more information to debugging output for retry time not reached.
3159 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3160 operations in malware.c.
3162 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3165 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3166 syslog_duplication set false, log lines that would normally be written
3167 both the the main log and to the reject log were not written to syslog at
3170 PH/14 In the default configuration, change the use of "message" in ACL warn
3171 statements to "add_header".
3173 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3174 not followed by a command (e.g. "seen endif").
3176 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3177 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3180 PH/17 Added extra conditions to the default value of delay_warning_condition
3184 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3185 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3186 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3189 The Auto-Submitted: and various List- headers are standardised, whereas I
3190 don't think Precedence: ever was.
3192 PH/18 Refactored debugging code in route_finduser() to show more information,
3193 in particular, the error code if getpwnam() issues one.
3195 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3196 This is apparently needed in addition to the PH/07 change above to avoid
3197 any possible encoding problems.
3199 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3200 but not after initializing Perl.
3202 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3203 output them only if debugging. By default they are written stderr,
3204 apparently, which is not desirable.
3206 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3209 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3212 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3214 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3215 authenticated or an ident call has been made. Suppress the default
3216 values for $authenticated_id and $authenticated_sender (but permit -oMai
3217 and -oMas) when testing with -bh.
3219 PH/24 Re-jigged the order of the tests in the default configuration so that the
3220 tests for valid domains and recipients precede the DNS black list and CSA
3221 tests, on the grounds that those ones are more expensive.
3223 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3224 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3225 This bug exists in every version of Exim that I still have, right back to
3228 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3229 However, an attempt to turn on case-sensitivity in a regex key by
3230 including (?-i) didn't work because the subject string was already
3231 lowercased, and the effects were non-intuitive. It turns out that a
3232 one-line patch can be used to allow (?-i) to work as expected.
3238 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3239 other effects) broke the use of negated acl sub-conditions.
3241 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3244 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3245 "Deny" causes Exim to reject the incoming connection with a 554 error.
3246 Unfortunately, if there is a major crisis, such as a disk failure,
3247 tcp-wrappers gives "deny", whereas what one would like would be some
3248 kind of temporary error. A kludge has been added to help with this.
3249 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3250 554 error is used if errno is still zero or contains ENOENT (which occurs
3251 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3254 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3256 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3257 errors. Otherwise a message that provokes a temporary error (when other
3258 messages do not) can cause a whole host to time out.
3260 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3261 addresses were routed directly to files or pipes from a redirect router.
3262 File deliveries just didn't batch; pipe deliveries might have suffered
3265 PH/06 A failure to get a lock for a hints database would erroneously always say
3266 "Failed to get write lock", even when it was really a read lock.
3268 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3269 of 0600. This has been changed to use the value of the lockfile_mode
3270 option (which defaults to 0600).
3272 PH/08 Applied small patch from the Sieve maintainer.
3274 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3275 folder from quota calculations, a direct delivery into this folder messed
3276 up the contents of the maildirsize file. This was because the regex was
3277 used only to exclude .Trash (or whatever) when the size of the mailbox
3278 was calculated. There was no check that a delivery was happening into an
3279 excluded directory. This bug has been fixed by ignoring all quota
3280 processing for deliveries into excluded directories.
3282 PH/10 Added the maildirfolder_create_regex option to appendfile.
3288 PH/01 The code for finding all the local interface addresses on a FreeBSD
3289 system running IPv6 was broken. This may well have applied to all BSD
3290 systems, as well as to others that have similar system calls. The broken
3291 code found IPv4 interfaces correctly, but gave incorrect values for the
3292 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3293 that it would not match correctly against @[] and not recognize the IPv6
3296 PH/02 The ipliteral router was not recognizing addresses of the form user@
3297 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3299 PH/03 Added disable_ipv6.
3301 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3302 decoded form, because it is most often used to construct To: headers
3303 lines in autoreplies, and the decoded form may well be syntactically
3304 invalid. However, $reply_address has leading white space removed, and all
3305 newlines turned into spaces so that the autoreply transport does not
3308 PH/05 If group was specified without a user on a router, and no group or user
3309 was specified on a transport, the group from the router was ignored.
3311 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3312 for visible compile-time settings that can be used to change these
3313 numbers, for those that want even more. Backwards compatibility with old
3314 spool files has been maintained. However, going back to a previous Exim
3315 release will lost any variables that are in spool files.
3317 PH/07 Two small changes when running in the test harness: increase delay when
3318 passing a TCP/IP connection to a new process, in case the original
3319 process has to generate a bounce, and remove special handling of
3320 127.0.0.2 (sic), which is no longer necessary.
3322 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3323 be the same on different OS.
3325 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3328 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3329 whether --show-vars was specified or not
3331 JJ/02 exipick: Added support for new ACL variable spool format introduced
3334 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3335 syntactically invalid From: or Reply-to: line, and a filter used this to
3336 generate an autoreply, and therefore failed to obtain an address for the
3337 autoreply, Exim could try to deliver to a non-existent relative file
3338 name, causing unrelated and misleading errors. What now happens is that
3339 it logs this as a hard delivery error, but does not attempt to create a
3342 PH/11 The exinext utility has a -C option for testing purposes, but although
3343 the given file was scanned by exinext itself; it wasn't being passed on
3344 when Exim was called.
3346 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3347 an end-of-file indication when reading a command response.
3349 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3350 compiled. In many other places in Exim, IPv6 addresses are always
3351 recognized, so I have changed this. It also means that IPv4 domain
3352 literals of the form [IPV4:n.n.n.n] are now always recognized.
3354 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3355 used if the router is not running as root, for example, when verifying at
3356 ACL time, or when using -bh. The debugging output from this situation was
3357 non-existent - all you got was a failure to exec. I have made two
3360 (a) Failures to set uid/gid, the current directory, or a process leader
3361 in a subprocess such as that created by queryprogram now generate
3362 suitable debugging output when -d is set.
3364 (b) The queryprogram router detects when it is not running as root,
3365 outputs suitable debugging information if -d is set, and then runs
3366 the subprocess without attempting to change uid/gid.
3368 PH/15 Minor change to Makefile for building test_host (undocumented testing
3371 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3372 additional section of a DNS packet that returns MX or SRV records.
3373 Instead, it always explicitly searches for A/AAAA records. This avoids
3374 major problems that occur when a DNS server includes only records of one
3375 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3376 fixed another bug: if SRV records were looked up and the corresponding
3377 address records were *not* found in the additional section, the port
3378 values from the SRV records were lost.
3380 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3381 using the correct key (the original address) when searching the retry
3382 rules in order to find which one to use for generating the retry hint.
3384 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3385 adding the default one. Similarly, if it contains a Reply-To: header, the
3386 errors_reply_to option, if set, is not used.
3388 PH/19 When calculating a retry time, Exim used to measure the "time since
3389 failure" by looking at the "first failed" field in the retry record. Now
3390 it does not use this if it is later than than the arrival time of the
3391 message. Instead it uses the arrival time. This makes for better
3392 behaviour in cases where some deliveries succeed, thus re-setting the
3393 "first failed" field. An example is a quota failure for a huge message
3394 when small messages continue to be delivered. Without this change, the
3395 "time since failure" will always be short, possible causing more frequent
3396 delivery attempts for the huge message than are intended.
3397 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3399 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3400 $1, $2, $3) because the numerical variables can be reset during some
3401 expansion items (e.g. "match"), thereby losing the authentication data.
3403 PH/21 Make -bV show the size of off_t variables so that the test suite can
3404 decide whether to run tests for quotas > 2G.
3406 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3407 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3408 filecount value is greater than 2G or if a quota value is greater than 2G
3409 on a system where the size of off_t is not greater than 4, a panic error
3412 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3413 never match. The debug and -bh output now contains an explicit error
3414 message indicating a malformed IPv4 address or mask.
3416 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3417 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3418 PH/23 above applies.
3420 PH/25 Do not write to syslog when running in the test harness. The only
3421 occasion when this arises is a failure to open the main or panic logs
3422 (for which there is an explicit test).
3424 PH/26 Added the /no_tell option to "control=freeze".
3426 PH/27 If a host name lookup failed very early in a connection, for example, if
3427 the IP address matched host_lookup and the reverse lookup yielded a name
3428 that did not have a forward lookup, an error message of the form "no IP
3429 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3430 could be logged. Now it outputs the IP address instead of "NULL".
3432 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3433 allows the sender and the authenticated sender to be set when
3434 submitting a message from within Exim. Since child_open_exim() is
3435 documented for local_scan(), the new function should be too.
3437 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3438 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3439 results in an empty string is now treated as unset.
3441 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3443 PH/31 Added sender_verify_fail logging option.
3445 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3446 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3447 tidied the source and removed it altogether.
3449 PH/33 When a queue run was abandoned because the load average was too high, a
3450 log line was always written; now it is written only if the queue_run log
3451 selector is set. In addition, the log line for abandonment now contains
3452 information about the queue run such as the pid. This is always present
3453 in "start" and "stop" lines but was omitted from the "abandon" line.
3455 PH/34 Omit spaces between a header name and the colon in the error message that
3456 is given when verify = headers_syntax fails (if there are lots of them,
3457 the message gets confusing).
3459 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3460 names, as there are now some PTR records that contain slashes. This check
3461 is only to protect against broken name servers that fall over on strange
3462 characters, so the fact that it applies to all lookups doesn't matter.
3464 PH/36 Now that the new test suite is complete, we can remove some of the
3465 special code in Exim that was needed for the old test suite. For example,
3466 sorting DNS records because real resolvers return them in an arbitrary
3467 order. The new test suite's fake resolver always returns records in the
3470 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3471 bounces) except when queue_only is set, to avoid logging races between
3472 the different processes.
3474 PH/38 Panic-die if .include specifies a non-absolute path.
3476 PH/39 A tweak to the "H" retry rule from its user.
3478 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3479 a label. They prevented compilation on older perls.
3481 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3482 a warning to be raised on newish perls.
3484 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3485 on queue. Changes to match documented behaviour of showing count of
3486 messages matching specified criteria.
3488 PH/40 Changed the default ident timeout from 30s to 5s.
3490 PH/41 Added support for the use of login_cap features, on those BSD systems
3491 that have them, for controlling the resources used by pipe deliveries.
3493 PH/42 The content-scanning code uses fopen() to create files in which to put
3494 message data. Previously it was not paying any attention to the mode of
3495 the files. Exim runs with umask(0) because the rest of the code creates
3496 files with open(), and sets the required mode explicitly. Thus, these
3497 files were ending up world-writeable. This was not a big issue, because,
3498 being within the spool directory, they were not world-accessible. I have
3499 created a function called modefopen, which takes an additional mode
3500 argument. It sets umask(777), creates the file, chmods it to the required
3501 mode, then resets the umask. All the relevant calls to fopen() in the
3502 content scanning code have been changed to use this function.
3504 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3505 to 24 hours. This avoids potential overflow problems when processing G
3506 and H retry rules. I suspect nobody ever tinkers with this value.
3508 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3510 PH/45 When the plaintext authenticator is running as a client, the server's
3511 challenges are checked to ensure they are valid base64 strings. By
3512 default, the authentication attempt is cancelled if an invalid string is
3513 received. Setting client_ignore_invalid_base64 true ignores these errors.
3514 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3515 they are received. Thus, the responses can be made to depend on the
3516 challenges. If an invalid string is ignored, an empty string is placed in
3519 PH/46 Messages that are created by the autoreply transport now contains a
3520 References: header, in accordance with RFCs 2822 and 3834.
3522 PH/47 Added authenticated_sender_force to the smtp transport.
3524 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3526 PH/49 Installed latest patch from the Sieve maintainer.
3528 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3529 was also set, the appendfile transport was unnecessarily scanning a
3530 directory of message files (e.g. for maildir delivery) to find the count
3531 of files (along with the size), even though it did not need this
3532 information. It now does the scan only if it needs to find either the
3533 size of the count of files.
3535 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3537 PH/52 Two bugs concerned with error handling when the smtp transport is
3540 (i) Exim was not creating retry information for temporary errors given
3541 for individual recipients after the DATA command when the smtp transport
3542 was used in LMTP mode. This meant that they could be retried too
3543 frequently, and not timed out correctly.
3545 (ii) Exim was setting the flag that allows error details to be returned
3546 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3547 recipients that were returned after the DATA command.
3549 PH/53 This is related to PH/52, but is more general: for any failing address,
3550 when detailed error information was permitted to be returned to the
3551 sender, but the error was temporary, then after the final timeout, only
3552 "retry timeout exceeded" was returned. Now it returns the full error as
3553 well as "retry timeout exceeded".
3555 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3556 do this, and (what is worse) MTAs that accept it.
3558 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3559 will now be deprecated.
3561 PH/56 New os.c-cygwin from the Cygwin maintainer.
3563 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3564 formats (previously only available in exim formats via -bpr, -bpru,
3565 and -bpra. Now also available in native and exiqgrep formats)
3567 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3568 with very large, slow to parse queues
3570 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3572 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3574 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3575 responses to authentication challenges, though it was showing the
3576 challenges; (ii) I've removed the CR characters from the debug output for
3579 PH/58 Allow for the insertion of a newline as well as a space when a string
3580 is turned into more than one encoded-word during RFC 2047 encoding. The
3581 Sieve code now uses this.
3583 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3584 data_4xx, lost_connection, tls_required.
3586 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3587 message was being sent as an SMTP response.
3589 PH/61 Add -l and -k options to exicyclog.
3591 PH/62 When verifying, if an address was redirected to one new address, so that
3592 verification continued, and the new address failed or deferred after
3593 having set something in $address_data, the value of $address_data was not
3594 passed back to the ACL. This was different to the case when no
3595 redirection occurred. The value is now passed back in both cases.
3597 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3598 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3599 use login_cap.h, so on its own it isn't the distinguishing feature. The
3600 new name refers directly to the setclassresources() function.
3602 PH/65 Added configuration files for NetBSD3.
3604 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3606 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3607 is preferred over IPv4.
3609 PH/68 The bounce_return_message and bounce_return_body options were not being
3610 honoured for bounces generated during the reception of non-SMTP messages.
3611 In particular, this applied to messages rejected by the ACL. This bug has
3612 been fixed. However, if bounce_return_message is true and bounce_return_
3613 body is false, the headers that are returned for a non-SMTP message
3614 include only those that have been read before the error was detected.
3615 (In the case of an ACL rejection, they have all been read.)
3617 PH/69 The HTML version of the specification is now built in a directory called
3618 spec_html instead of spec.html, because the latter looks like a path with
3619 a MIME-type, and this confuses some software.
3621 PH/70 Catch two compiler warnings in sieve.c.
3623 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3624 function verify_get_ident() calls ip_connect() to connect a socket, but
3625 if the "connect()" function timed out, ip_connect() used to close the
3626 socket. However, verify_get_ident() also closes the socket later, and in
3627 between Exim writes to the log, which may get opened at this point. When
3628 the socket was closed in ip_connect(), the log could get the same file
3629 descriptor number as the socket. This naturally causes chaos. The fix is
3630 not to close the socket in ip_connect(); the socket should be closed by
3631 the function that creates it. There was only one place in the code where
3632 this was missing, in the iplookup router, which I don't think anybody now
3633 uses, but I've fixed it anyway.
3635 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3636 well as to direct DNS lookups. Otherwise the handling of names in host
3637 lists is inconsistent and therefore confusing.
3643 PH/01 Two changes to the default runtime configuration:
3645 (1) Move the checks for relay_from_hosts and authenticated clients from
3646 after to before the (commented out) DNS black list checks.
3648 (2) Add control=submission to the relay_from_hosts and authenticated
3649 clients checks, on the grounds that messages accepted by these
3650 statements are most likely to be submissions.
3652 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3654 (1) Generate an error if the third argument for the ${prvs expansion is
3657 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3660 (3) Reset the variables that are obtained from the first argument of
3661 ${prvscheck and used in the second argument before leaving the code,
3662 because their memory is reclaimed, so using them afterwards may do
3665 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3666 one (it's much easier than Tom thought :-).
3668 (5) Because of (4), we can now allow for the use of $prvscheck_result
3669 inside the third argument.
3671 PH/03 For some reason, the default setting of PATH when running a command from
3672 a pipe transport was just "/usr/bin". I have changed it to
3675 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3676 anything to be listed in the output from -bV.
3678 PH/05 When a filter generated an autoreply, the entire To: header line was
3679 quoted in the delivery log line, like this:
3681 => >A.N.Other <ano@some.domain> <original@ddress> ...
3683 This has been changed so that it extracts the operative address. There
3684 may be more than one such address. If so, they are comma-separated, like
3687 => >ano@some.domain,ona@other.domain <original@ddress> ...
3689 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3690 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3691 not being looked up in the rDNS to get a host name, Exim was showing the
3692 IP address twice in Received: lines, even though the IP addresses were
3693 identical. For example:
3695 Received: from [1.2.3.4] (helo=[1.2.3.4])
3697 However, if the real host name was known, it was omitting the HELO data
3698 if it matched the actual IP address. This has been tidied up so that it
3699 doesn't show the same IP address twice.
3701 PH/07 When both +timestamp and +memory debugging was on, the value given by
3702 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3703 called by the debug printing, thereby overwriting the timestamp buffer.
3704 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3706 PH/08 When the original message was included in an autoreply transport, it
3707 always said "this is a copy of the message, including all the headers",
3708 even if body_only or headers_only was set. It now gives an appropriate
3711 PH/09 Applied a patch from the Sieve maintainer which:
3713 o fixes some comments
3714 o adds the (disabled) notify extension core
3715 o adds some debug output for the result of if/elsif tests
3716 o points to the current vacation draft in the documentation
3717 and documents the missing references header update
3721 o fixes a bug in processing the envelope test (when testing
3722 multiple envelope elements, the last element determined the
3725 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3726 Electronic Mail") by including:
3728 Auto-submitted: auto-generated
3730 in the messages that it generates (bounce messages and others, such as
3731 warnings). In the case of bounce messages for non-SMTP messages, there was
3732 also a typo: it was using "Auto_submitted" (underscore instead of
3733 hyphen). Since every message generated by Exim is necessarily in response
3734 to another message, thes have all been changed to:
3736 Auto-Submitted: auto-replied
3738 in accordance with these statements in the RFC:
3740 The auto-replied keyword:
3742 - SHOULD be used on messages sent in direct response to another
3743 message by an automatic process,
3745 - MUST NOT be used on manually-generated messages,
3747 - MAY be used on Delivery Status Notifications (DSNs) and Message
3748 Disposition Notifications (MDNs),
3750 - MUST NOT be used on messages generated by automatic or periodic
3751 processes, except for messages which are automatic responses to
3754 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3755 to the default Received: header definition.
3757 PH/12 Added log selector acl_warn_skipped (default on).
3759 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3760 variables because (a) they are in the wrong storage pool and (b) even if
3761 they were copied, it wouldn't work properly because of the caching.
3763 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3764 checking when decoding. Apparently there are clients that generate
3765 overlong encoded strings. Why am I not surprised?
3767 PH/15 If the first argument of "${if match_address" was not empty, but did not
3768 contain an "@" character, Exim crashed. Now it writes a panic log message
3769 and treats the condition as false.
3771 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3773 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3774 extension "envelope-auth". The code is finished and in agreement with
3775 other implementations, but there is no documentation so far and in fact,
3776 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3777 not changing the active code.
3779 Print executed "if" and "elsif" statements when debugging is used. This
3780 helps a great deal to understand what a filter does.
3782 Document more things not specified clearly in RFC3028. I had all this
3783 sorted out, when out of a sudden new issues came to my mind. Oops."
3785 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3788 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3789 canonical form (as documented). However, after a host name lookup from
3790 the IP address, check_host() was doing a simple string comparison with
3791 addresses acquired from the DNS when checking that the found name did
3792 have the original IP as one of its addresses. Since any found IPv6
3793 addresses are likely to be in abbreviated form, the comparison could
3794 fail. Luckily, there already exists a function for doing the comparison
3795 by converting both addresses to binary, so now that is used instead of
3796 the text comparison.
3798 PH/20 There was another similar case to PH/19, when a complete host name was
3799 given in a host list; looking up its IP address could give an abbreviated
3800 form, whereas the current host's name might or might not be abbreviated.
3801 The same fix has been applied.
3807 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3808 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3811 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3812 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3814 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3816 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3817 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3818 Also, for "Auto-Submitted", treat anything other than "no" as
3819 non-personal, in accordance with RFC 3834. (Previously it treated
3820 anything starting "auto-" as non-personal.)
3822 TF/01 The control=submission/name=... option had a problem with syntax
3823 errors if the name included a slash character. The /name= option
3824 now slurps the rest of the string, so it can include any characters
3825 but it must come last in the list of options (after /sender_retain
3828 PH/05 Some modifications to the interface to the fake nameserver for the new
3836 TK/01 Added the "success_on_redirect" address verification option. See
3837 NewStuff for rationale and an example.
3839 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3841 PH/02 Patch to exigrep to allow it to work on syslog lines.
3843 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3844 fread() to skip over the body file's header line, because in Cygwin the
3845 header line is locked and is inaccessible.
3847 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3848 co-exist for some time) to make it clear that it is the Exim ID that is
3849 referenced, not the Message-ID: header line.
3851 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3852 string_format() function, because snprintf() does not exist on all
3855 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3856 ${stat: expansion item.
3858 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3859 protocol synchronization error", to keep the pedants happy.
3861 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3862 well as for IRIX systems, when gcc is being used. See the host.c source
3865 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3867 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3870 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3871 transport and to the smtp transport in LMTP mode.
3873 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3875 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3876 run a filter in a subprocess. This could lead to confusion in subsequent
3877 lookups in the parent process. There should also be a search_tidyup() at
3878 the end of the subprocess.
3880 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3881 only if the host matched helo_try_verify_hosts, which caused the
3882 verification to occur when the EHLO/HELO command was issued. The ACL just
3883 tested the remembered result. Now, if a previous verification attempt has
3884 not happened, "verify = helo" does it there and then.
3886 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3888 TK/03 Fix log output including CR from clamd.
3890 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3891 exist provoked a memory error which could cause a segfault.
3893 PH/15 Installed PCRE 6.2
3895 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3897 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3898 of the problem. Specifically, suggested +O2 rather than +O1 for the
3901 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3903 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3904 record, the port was not correctly logged when the outgoing_port log
3905 selector was set (it logged the transort's default port).
3907 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3908 fallback_hosts, and "hosts" in the smtp transport.
3910 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3911 host errors such as "Connection refused".
3913 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3914 authentication with radiusclient 0.4.9:
3916 - Error returned from rc_read_config was caught wrongly
3917 - Username/password not passed on to radius server due to wrong length.
3919 The presumption is that some radiusclient API changes for 4.51/PH/17
3920 were not taken care of correctly. The code is still untested by me (my
3921 Linux distribution still has 0.3.2 of radiusclient), but it was
3922 contributed by a Radius user.
3924 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3925 expanding the "port" option of the smtp transport.
3927 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3928 while reading a MIME header. Thanks to Tom Hughes for a patch.
3930 PH/24 Include config.h inside local_scan.h so that configuration settings are
3933 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3934 that in an ACL for RCPT (for example), you can examine exactly what was
3937 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3938 commands, but it was not correctly comparing the address with the actual
3939 client host address. Thus, it would show the EHLO address in Received:
3940 header lines when this was not necessary.
3942 PH/27 Added the % operator to ${eval:}.
3944 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3945 it should be ignoring failures (because with -C, for example, it has lost
3946 privilege). It wasn't ignoring creation failures other than "already
3949 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3952 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3953 because an input line was too long, either on its own, or by virtue of
3954 too many continuations, the temporary file was not being removed, and the
3955 return code was incorrect.
3957 PH/31 Missing "BOOL" in function definition in filtertest.c.
3959 PH/32 Applied Sieve patches from the maintainer.
3961 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3963 PH/33 Added "verify = not_blind".
3965 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3966 Local/Makefile (with some defaults set). These are used in built scripts
3967 such as exicyclog, but they have never been used in the exim_install
3968 script (though there are many overriding facilities there). I have
3969 arranged that the exim_install script now takes note of these two
3972 PH/35 Installed configuration files for Dragonfly.
3974 PH/36 When a locally submitted message by a trusted user did not contain a
3975 From: header, and the sender address was obtained from -f or from an SMTP
3976 MAIL command, and the trusted user did not use -F to supply a sender
3977 name, $originator_name was incorrectly used when constructing a From:
3978 header. Furthermore, $originator_name was used for submission mode
3979 messages from external hosts without From: headers in a similar way,
3980 which is clearly wrong.
3982 PH/37 Added control=suppress_local_fixups.
3984 PH/38 When log_selector = +received_sender was set, and the addition of the
3985 sender made the log line's construction buffer exactly full, or one byte
3986 less than full, an overflow happened when the terminating "\n" was
3989 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3990 when the result of a list match is failure because a DNS lookup failed.
3992 PH/40 RM_COMMAND is now used in the building process.
3994 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3995 the "build-* directories that it finds.
3997 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3998 address in a domain literal was a prefix of an interface address.
4000 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4001 when verifying a sender address, unless rewrite_headers is false.
4003 PH/44 Wrote a long comment about why errors_to addresses are verified as
4004 recipients, not senders.
4006 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4007 the ratelimit ACL was added.
4009 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4011 PH/46 Added extra information about PostgreSQL errors to the error string.
4013 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4014 avoiding the need to install special zones in a real server. This is
4015 backwards compatible; if it can't find the fake resolver, it drops back.
4016 Thus, both old and new test suites can be run.
4018 TF/02 Added util/ratelimit.pl
4020 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4021 clock is set back in time.
4023 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4024 Candler <B.Candler@pobox.com>.
4026 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4027 OK for addresses that are the result of redirections.
4029 PH/48 A number of further additions for the benefit of the new test suite,
4030 including a fake gethostbyname() that interfaces to the fake DNS resolver
4033 TF/06 The fix for widen_domains has also been applied to qualify_single and
4034 search_parents which are the other dnslookup options that can cause
4037 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4040 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4042 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4043 service" if it encountered an MX record with an empty target hostname.
4044 The message is now "an MX or SRV record indicated no SMTP service".
4046 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4047 if the DNS of the sending site is misconfigured. This is quite a
4048 common situation. This change restores the behaviour of treating a
4049 helo verification defer as a failure.
4051 PH/51 If self=fail was set on a router, the bounce message did not include the
4052 actual error message.
4058 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4060 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4061 command itself ended in error, the underlying message about the transport
4062 filter timeout was being overwritten with the pipe command error. Now the
4063 underlying error message should be appended to the second error message.
4065 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4067 PH/02 Reduce the amount of output that "make" produces by default. Full output
4068 can still be requested.
4070 PH/03 The warning log line about a condition test deferring for a "warn" verb
4071 was being output only once per connection, rather than after each
4072 occurrence (because it was using the same function as for successful
4073 "warn" verbs). This seems wrong, so I have changed it.
4075 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4076 it should not have, which might have caused a crash in the right
4077 circumstances, but probably never did.
4079 PH/04 Installed a modified version of Tony Finch's patch to make submission
4080 mode fix the return path as well as the Sender: header line, and to
4081 add a /name= option so that you can make the user's friendly name appear
4084 TF/03 Added the control = fakedefer ACL modifier.
4086 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4087 Mark Lowes for thorough testing.
4089 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4091 TK/03 Merged latest SRS patch from Miles Wilton.
4093 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4094 with the definition in sysexits.h (which is #included earlier).
4095 Fortunately, Exim does not actually use EX_OK. The code used to try to
4096 preserve the sysexits.h value, by assuming that macro definitions were
4097 scanned for macro replacements. I have been disabused of this notion,
4098 so now the code just undefines EX_OK before #including unistd.h.
4100 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4101 in the smtp transport. When a block could not be written in a single
4102 write() function, the timeout was being re-applied to each part-write.
4103 This seems wrong - if the receiver was accepting one byte at a time it
4104 would take for ever. The timeout is now adjusted when this happens. It
4105 doesn't have to be particularly precise.
4107 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4108 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4110 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4111 without a domain if the check to <postmaster@domain> fails.
4113 SC/01 Eximstats: added -xls and the ability to specify output files
4114 (patch written by Frank Heydlauf).
4116 SC/02 Eximstats: use FileHandles for outputting results.
4118 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4120 SC/04 Eximstats: fixed display of large numbers with -nvr option
4122 SC/05 Eximstats: fixed merging of reports with empty tables.
4124 SC/06 Eximstats: added the -include_original_destination flag
4126 SC/07 Eximstats: removed tabs and trailing whitespace.
4128 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4130 TK/06 MBOX spool code: Add real "From " MBOX separator line
4131 so the .eml file is really in mbox format (even though
4132 most programs do not really care). Patch from Alex Miller.
4134 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4135 The latter is generated from $received_to and is only set if the
4136 message has one envelope recipient. SA can use these headers,
4137 obviously out-of-the-box. Patch from Alex Miller.
4139 PH/08 The ${def test on a variable was returning false if the variable's
4140 value was "0", contrary to what the specification has always said!
4141 The result should be true unless the variable is empty.
4143 PH/09 The syntax error of a character other than { following "${if
4144 def:variable_name" (after optional whitespace) was not being diagnosed.
4145 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4146 accidental colon was present, for example, could give incorrect results.
4148 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4149 result is used (not including appendfile, where other changes are about
4152 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4153 This involved changing a lot of size variables from int to off_t. It
4154 should work with maildirs and everything.
4156 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4157 spamd dying while we are connected to it.
4159 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4162 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4163 function for BDB 4.3.
4165 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4167 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4168 the sources for PCRE have been re-arranged and more files are now
4171 PH/15 The code I had for printing potentially long long variables in PH/11
4172 above was not the best (it lost precision). The length of off_t variables
4173 is now inspected at build time, and an appropriate printing format (%ld
4174 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4175 to be "long long int" or "long int". This is needed for the internal
4176 formatting function string_vformat().
4178 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4179 the configuration file to be ":syslog", then the script "guesses" where
4180 the logs files are, rather than using the compiled in default. In our
4181 case the guess is not the same as the compiled default, so the script
4182 suddenly stopped working when I started to use syslog. The patch checks
4183 to see if log_file_path is "". If so, it attempts to read it from exim
4184 with no configuration file to get the compiled in version, before it
4185 falls back to the previous guessing code."
4187 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4188 implementing BATV in an Exim configuration. See NewStuff for the gory
4191 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4192 Makefile that are specific to HP-UX.
4194 PH/18 If the "use_postmaster" option was set for a recipient callout together
4195 with the "random" option, the postmaster address was used as the MAIL
4196 FROM address for the random test, but not for the subsequent recipient
4197 test. It is now used for both.
4199 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4200 patch removes a few documentation additions to RFC 3028, because the
4201 latest draft now contains them. It adds the new en;ascii-case comparator
4202 and a new error check for 8bit text in MIME parts. Comparator and
4203 require names are now matched exactly. I enabled the subaddress
4204 extension, but it is not well tested yet (read: it works for me)."
4206 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4207 rework some of the code of TK/09 above to avoid the hardwired use of
4208 "%lld" and "long long". Replaced the call to snprintf() with a call to
4211 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4212 records point to non-existent hosts", "retry timeout exceeded", and
4213 "retry time not reached for any host after a long failure period".
4215 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4216 experimental DomainKeys support:
4218 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4219 (2) On an error such as an illegally used "control", the wrong name for
4220 the control was given.
4222 These problems did NOT occur unless DomainKeys support was compiled.
4224 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4226 PH/24 Added ${if match_ip condition.
4228 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4229 databases so that it will be absolutely obvious if a crash occurs in the
4230 DB library. This is a regular occurrence (often caused by mis-matched
4233 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4234 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4235 picked up on a user's system that detects such things. There doesn't seem
4236 to be a gcc warning option for this - only an attribute that has to be
4237 put on the function's prototype. It seems that in Fedora Core 4 they have
4238 set this on a number of new functions. No doubt there will be more in due
4241 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4242 specify a transport. However, if an address that was verified by such a
4243 router was the subject of a callout, Exim crashed because it tried to
4244 read the rcpt_include_affixes from the non-existent transport. Now it
4245 just assumes that the setting of that option is false. This bug was
4246 introduced by 4.51/PH/31.
4248 PH/28 Changed -d+all to exclude +memory, because that information is very
4249 rarely of interest, but it makes the output a lot bigger. People tend to
4250 do -d+all out of habit.
4252 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4253 code in os-type was giving problems when libc.so lives in lib64, like on
4256 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4257 aren't the modern standard, and it seems that some systems' include files
4258 don't always have them. Exim was already checking for some of the newer
4259 ones like T_AAAA, and defining it itself. I've added checks for all the
4260 record types that Exim uses.
4262 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4263 not automatically generating a new one, as it is supposed to. This
4264 prevented TLS from working. If the file did exist, but contained invalid
4265 data, a new version was generated, as expected. It was only the case of a
4266 non-existent file that was broken.
4268 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4269 with a change in libdomainkeys > 0.64.
4271 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4272 from DNS. If the selector record carries the flag, it now has
4273 precedence over the domain-wide flag.
4275 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4277 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4278 the use of a port name that isn't defined in /etc/services) occurred, the
4279 message was deferred as in a normal delivery, and thus remained on the
4280 spool, instead of being failed because of the mua_wrapper setting. This
4281 is now fixed, and I tidied up some of the mua_wrapper messages at the
4284 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4285 the messages in a hash of arrays rather than using individual hashes.
4286 This is a bit cleaner and results in dramatic memory savings, albeit
4287 at a slight CPU cost.
4289 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4290 as requested by Marc Sherman.
4292 SC/10 Eximstats: added histograms for user specified patterns as requested
4295 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4297 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4298 fopen() in the content-scanning modules that did not already have it.
4304 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4305 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4307 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4309 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4311 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4312 to test Sieve filters that use "vacation".
4314 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4315 that changes the way the GnuTLS parameters are stored in the cache file.
4316 The new format can be generated externally. For backward compatibility,
4317 if the data in the cache doesn't make sense, Exim assumes it has read an
4318 old-format file, and it generates new data and writes a new file. This
4319 means that you can't go back to an older release without removing the
4322 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4323 work if there are any delivery delays because "one_time" forces the
4324 parent to be marked "delivered", so its unseen clone is never tried
4325 again. For this reason, Exim now forbids the simultaneous setting of
4328 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4329 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4330 entry if you want to know the details. The fix, however, neglected to
4331 consider the case when local delivery batching is involved. The test for
4332 "previously delivered" was not happening when checking to see if an
4333 address could be batched with a previous (undelivered) one; under
4334 certain circumstances this could lead to multiple deliveries to the same
4337 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4338 in its include files, and this causes problems building Exim.
4340 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4341 header_syntax) but Exim was just ignoring anything given after a slash.
4342 In particular, this caused confusion with an attempt to use "verify =
4343 reverse_host_lookup/defer_ok". An error is now given when options are
4344 supplied for verify items that do not have them. (Maybe reverse_host_
4345 lookup should have a defer_ok option, but that's a different point.)
4347 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4348 defined by RFC 821) to 2048, because there were problems with some AUTH
4349 commands, and RFC 1869 says the size should be increased for extended
4350 SMTP commands that take arguments.
4352 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4355 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4356 "unknown" error; now it says that the functionality isn't in the binary.
4358 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4359 an address' error message when a string expansion fails (syntax or
4360 whatever). Otherwise the password may appear in the log. Following change
4361 PH/42 below, there is no longer a chance of it appearing in a bounce
4364 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4366 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4367 of its addresses was ever tried. (Bugzilla bug #2.)
4369 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4370 the result incorrectly in the debug output. (It correctly added a newline
4371 to what was transported.)
4373 TF/01 Added $received_time.
4375 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4376 commented out examples of how to interface to a virus scanner and to
4377 SpamAssassin. Also added commented examples of av_scanner and
4378 spamd_address settings.
4380 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4381 and controls are allowed in which ACLs. There were a couple of minor
4382 errors. Some of the entries in the conditions table (which is a table of
4383 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4384 negation of where the condition IS allowed.
4386 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4388 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4389 header file does not have a version number, so I've had to invent a new
4390 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4391 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4392 radiusclient), but it was contributed by a Radius user.
4394 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4395 files or directories, for OpenSSL.
4397 PH/19 When an Exim process that is running as root has to create an Exim log
4398 file, it does so in a subprocess that runs as exim:exim so as to get the
4399 ownership right at creation (otherwise, other Exim processes might see
4400 the file with the wrong ownership). There was no test for failure of this
4401 fork() call, which would lead to the process getting stuck as it waited
4402 for a non-existent subprocess. Forks do occasionally fail when resources
4403 run out. I reviewed all the other calls to fork(); they all seem to check
4406 PH/20 When checking for unexpected SMTP input at connect time (before writing
4407 the banner), Exim was not dealing correctly with a non-positive return
4408 from the read() function. If the client had disconnected by this time,
4409 the result was a log entry for a synchronization error with an empty
4410 string after "input=" when read() returned zero. If read() returned -1
4411 (an event I could not check), uninitialized data bytes were printed.
4412 There were reports of junk text (parts of files, etc) appearing after
4415 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4417 PH/22 Added support for macro redefinition, and (re)definition in between
4418 driver and ACL definitions.
4420 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4421 forgetting to use the resulting value; it was using the unexpanded value.
4423 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4424 hadn't been configured. The fix is from Juergen Kreileder, who
4425 understands it better than I do:
4427 "Here's what I see happening with three configured cyrus_sasl
4428 authenticators configured (plain, login, cram-md5):
4430 On startup auth_cyrus_sasl_init() gets called for each of these.
4431 This means three calls to sasl_listmech() without a specified mech_list.
4432 => SASL tests which mechs of all available mechs actually work
4433 => three warnings about OTP not working
4434 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4436 With the patch, sasl_listmech() also gets called three times. But now
4437 SASL's mech_list option is set to the server_mech specified in the the
4438 authenticator. Or in other words, the answer from sasl_listmech()
4439 gets limited to just the mech you're testing for (which is different
4441 => the return list contains just 'plain' or 'login', 'cram-md5' or
4442 nothing depending on the value of ob->server_mech.
4444 I've just tested the patch: Authentication still works fine,
4445 unavailable mechs specified in the exim configuration are still
4446 caught, and the auth.log warnings about OTP are gone."
4448 PH/25 When debugging is enabled, the contents of the command line are added
4449 to the debugging output, even when log_selector=+arguments is not
4452 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4453 answer is "GNU", and only if the return is "GNU/something" is the answer
4456 PH/27 $acl_verify_message is now set immediately after the failure of a
4457 verification in an ACL, and so is available in subsequent modifiers. In
4458 particular, the message can be preserved by coding like this:
4460 warn !verify = sender
4461 set acl_m0 = $acl_verify_message
4463 Previously, $acl_verify_message was set only while expanding "message"
4464 and "log_message" when a very denied access.
4466 PH/28 Modified OS/os.c-Linux with
4468 -#ifndef OS_LOAD_AVERAGE
4469 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4471 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4472 nomenclature these days.)
4474 PH/29 Installed patch from the Sieve maintainer that adds the options
4475 sieve_useraddress and sieve_subaddress to the redirect router.
4477 PH/30 In these circumstances:
4478 . Two addresses routed to the same list of hosts;
4479 . First host does not offer TLS;
4480 . First host accepts first address;
4481 . First host gives temporary error to second address;
4482 . Second host offers TLS and a TLS session is established;
4483 . Second host accepts second address.
4484 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4485 and peerdn, if requested) that were in fact used only for the second
4488 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4489 attention to any local part prefix or suffix that was matched by the
4490 router that accepted the address. It now behaves in the same way as it
4491 does for delivery: the affixes are removed from the local part unless
4492 rcpt_include_affixes is set on the transport.
4494 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4495 timeout during the DATA phase of an incoming message.
4497 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4498 applied a patch sanctioned by the Sieve maintainer.
4500 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4501 the uid or gid is negative. A case of a negative gid caused this to be
4502 noticed. The fix allows for either to be negative.
4504 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4505 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4508 PH/36 The change PH/12 above was broken. Fixed it.
4510 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4511 the grounds that routing the same address twice would always produce the
4512 same answer. This might have been true once, but it is certainly no
4513 longer true now. Routing a child address may depend on the previous
4514 routing that produced that child. Some complicated redirection strategies
4515 went wrong when messages had multiple recipients, and made Exim's
4516 behaviour dependent on the order in which the addresses were given.
4518 I have moved the duplicate checking until after the routing is complete.
4519 Exim scans the addresses that are assigned to local and remote
4520 transports, and removes any duplicates. This means that more work will be
4521 done, as duplicates will always all be routed, but duplicates are
4522 presumably rare, so I don't expect this is of any significance.
4524 For deliveries to pipes, files, and autoreplies, the duplicate checking
4525 still happens during the routing process, since they are not going to be
4528 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4529 It corrects a timeout issue with spamd. This is Ian's comment: "The
4530 background is that sometimes spamd either never reads data from a
4531 connection it has accepted, or it never writes response data. The exiscan
4532 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4533 blindly assumes that writes won't block so it may never time out."
4535 PH/39 Allow G after quota size as well as K and M.
4537 PH/40 The value set for $authenticated_id in an authenticator may not contain
4538 binary zeroes or newlines because the value is written to log lines and
4539 to spool files. There was no check on this. Now the value is run through
4540 the string_printing() function so that such characters are converted to
4541 printable escape sequences.
4543 PH/41 $message_linecount is a new variable that contains the total number of
4544 lines in the message. Compare $body_linecount, which is the count for the
4547 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4548 bounce and delay warning messages, except in certain special cases, which
4551 (a) An SMTP error message from a remote host;
4552 (b) A message specified in a :fail: redirection;
4553 (c) A message specified in a "fail" command in a system filter;
4554 (d) A message specified in a FAIL return from the queryprogram router;
4555 (e) A message specified by the cannot_route_message router option.
4557 In these cases only, Exim does include the error details in bounce and
4558 warning messages. There are also a few cases where bland messages such
4559 as "unrouteable address" or "local delivery error" are given.
4561 PH/43 $value is now also set for the "else" part of a ${run expansion.
4563 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4564 being worked on, but at least Exim now implements the latest version to
4567 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4568 process to complete was treated as a delivery failure, a timeout while
4569 writing the message to the pipe was logged, but erroneously treated as a
4570 successful delivery. Such timeouts include transport filter timeouts. For
4571 consistency with the overall process timeout, these timeouts are now
4572 treated as errors, giving rise to delivery failures by default. However,
4573 there is now a new Boolean option for the pipe transport called
4574 timeout_defer, which, if set TRUE, converts the failures into defers for
4575 both kinds of timeout. A transport filter timeout is now identified in
4578 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4579 systems where "make" and "gmake" are different, calling "gmake" at top
4580 level broke things. I've arranged for the value of $(MAKE) to be passed
4581 from the Makefile to this script so that it can call the same version of
4585 A note about Exim versions 4.44 and 4.50
4586 ----------------------------------------
4588 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4589 changes of various kinds. As a consequence, a big documentation update was
4590 needed. This delayed the release for rather longer than seemed good, especially
4591 in the light of a couple of (minor) security issues. Therefore, the changes
4592 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4593 release. So 4.44 and 4.50 are in effect two different branches that both start
4596 I have left the 4.50 change log unchanged; it contains all the changes since
4597 4.43. The change log for 4.44 is below; many of its items are identical to
4598 those for 4.50. This seems to be the most sensible way to preserve the
4599 historical information.
4605 1. Minor wording change to the doc/README.SIEVE file.
4607 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4608 computation of the current number of files was incorrect.
4610 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4611 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4614 4. Give more explanation in the error message when the command for a transport
4615 filter fails to execute.
4617 5. There are several places where Exim runs a non-Exim command in a
4618 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4619 was being done only for the command run by the queryprogram router. It is
4620 now done for all such subprocesses. The other cases are: ${run, transport
4621 filters, and the commands run by the lmtp and pipe transports.
4623 6. Added CONFIGURE_GROUP build-time option.
4625 7. Some older OS have a limit of 256 on the maximum number of file
4626 descriptors. Exim was using setrlimit() to set 1000 as a large value
4627 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4628 systems. I've change it so that if it can't get 1000, it tries for 256.
4630 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4631 was an oversight, and furthermore, ever since the addition of extra
4632 controls (e.g. 4.43/32), the checks on when to allow different forms of
4633 "control" were broken. There should now be diagnostics for all cases when a
4634 control that does not make sense is encountered.
4636 9. Added the /retain_sender option to "control=submission".
4638 10. $recipients is now available in the predata ACL (oversight).
4640 11. Tidy the search cache before the fork to do a delivery from a message
4641 received from the command line. Otherwise the child will trigger a lookup
4642 failure and thereby defer the delivery if it tries to use (for example) a
4643 cached ldap connection that the parent has called unbind on.
4645 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4646 of $address_data from the recipient verification was clobbered by the
4647 sender verification.
4649 13. The value of address_data from a sender verification is now available in
4650 $sender_address_data in subsequent conditions in the ACL statement.
4652 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4654 15. Added a new option "connect=<time>" to callout options, to set a different
4657 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4658 was its contents. (It was OK if the option was not defined at all.)
4660 17. A "Completed" log line is now written for messages that are removed from
4661 the spool by the -Mrm option.
4663 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4664 information about exactly what failed.
4666 19. Added -dd to debug only the daemon process.
4668 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4669 handles timeouts, both on the server side and network timeouts. Renamed the
4670 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4672 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4673 It is now set to "smtps".
4675 22. $host_address is now set to the target address during the checking of
4676 ignore_target_hosts.
4678 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4679 being passed; this would have caused $sender_host_name to have been used if
4680 matching the list had actually called for a host name (not very likely,
4681 since this list is usually IP addresses). A host name is now passed as
4684 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4685 code that specifies a non-restarting handler (typically sigaction() in
4686 modern systems) in an attempt to fix a rare and obscure crash bug.
4688 25. Narrowed the window for a race in the daemon that could cause it to ignore
4689 SIGCHLD signals. This is not a major problem, because they are used only to
4690 wake it up if nothing else does.
4692 26. A malformed maildirsize file could cause Exim to calculate negative values
4693 for the mailbox size or file count. Odd effects could occur as a result.
4694 The maildirsize information is now recalculated if the size or filecount
4697 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4698 support for a long time. Removed HAVE_SYS_VFS_H.
4700 28. Installed the latest version of exipick from John Jetmore.
4702 29. In an address list, if the pattern was not a regular expression, an empty
4703 subject address (from a bounce message) matched only if the pattern was an
4704 empty string. Non-empty patterns were not even tested. This was the wrong
4705 because it is perfectly reasonable to use an empty address as part of a
4706 database query. An empty address is now tested by patterns that are
4707 lookups. However, all the other forms of pattern expect the subject to
4708 contain a local part and a domain, and therefore, for them, an empty
4709 address still always fails if the pattern is not itself empty.
4711 30. Exim went into a mad DNS loop when attempting to do a callout where the
4712 host was specified on an smtp transport, and looking it up yielded more
4713 than one IP address.
4715 31. Re-factored the code for checking spool and log partition space into a
4716 function that finds that data and another that does the check. The former
4717 is then used to implement four new variables: $spool_space, $log_space,
4718 $spool_inodes, and $log_inodes.
4720 32. The RFC2047 encoding function was originally intended for short strings
4721 such as real names; it was not keeping to the 75-character limit for
4722 encoded words that the RFC imposes. It now respects the limit, and
4723 generates multiple encoded words if necessary. To be on the safe side, I
4724 have increased the buffer size for the ${rfc2047: expansion operator from
4727 33. It is now permitted to omit both strings after an "if" condition; if the
4728 condition is true, the result is "true". As before, when the second string
4729 is omitted, a false condition yields an empty string. This makes it less
4730 cumbersome to write custom ACL and router conditions.
4732 34. Failure to deliver a bounce message always caused it to be frozen, even if
4733 there was an errors_to setting on the router. The errors_to setting is now
4736 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4737 canonical form (fully expanded) before being placed in
4738 $sender_host_address.
4740 36. The table in the code that translates DNS record types into text (T_A to
4741 "A" for instance) was missing entries for NS and CNAME. It is just possible
4742 that this could have caused confusion if both these types were looked up
4743 for the same domain, because the text type is used as part of Exim's
4744 per-process caching. But the chance of anyone hitting this buglet seems
4747 37. The dnsdb lookup has been extended in a number of ways.
4749 (1) There is a new type, "zns", which walks up the domain tree until it
4750 finds some nameserver records. It should be used with care.
4752 (2) There is a new type, "mxh", which is like "mx" except that it returns
4753 just the host names, not the priorities.
4755 (3) It is now possible to give a list of domains (or IP addresses) to be
4756 looked up. The behaviour when one of the lookups defers can be
4757 controlled by a keyword.
4759 (4) It is now possible to specify the separator character for use when
4760 multiple records are returned.
4762 38. The dnslists ACL condition has been extended: it is now possible to supply
4763 a list of IP addresses and/or domains to be looked up in a particular DNS
4766 39. Added log_selector=+queue_time_overall.
4768 40. When running the queue in the test harness, wait just a tad after forking a
4769 delivery process, to get repeatability of debugging output.
4771 41. Include certificate and key file names in error message when GnuTLS fails
4772 to set them up, because the GnuTLS error message doesn't include the name
4773 of the failing file when there is a problem reading it.
4775 42. Allow both -bf and -bF in the same test run.
4777 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4779 44. The "Exiscan patch" is now merged into the mainline Exim source.
4781 45. Sometimes the final signoff response after QUIT could fail to get
4782 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4783 < 0 before doing a fflush(). This bug looks as though it goes back to the
4784 introduction of TLS in release 3.20, but "sometimes" must have been rare
4785 because the tests only now provoked it.
4787 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4788 (this can affect the format of dates).
4790 47. exim_tidydb, when checking for the continued existence of a message for
4791 which it has found a message-specific retry record, was not finding
4792 messages that were in split spool directories. Consequently, it was
4793 deleting retry records that should have stayed in existence.
4795 48. Steve fixed some bugs in eximstats.
4797 49. The SPA authentication driver was not abandoning authentication and moving
4798 on to the next authenticator when an expansion was forced to fail,
4799 contradicting the general specification for all authenticators. Instead it
4800 was generating a temporary error. It now behaves as specified.
4802 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4803 (the order specifies the preference for clients). The order is now AES256,
4804 AES128, 3DES, ARCFOUR128.
4806 51. Small patch to Sieve code - explicitly set From: when generating an
4809 52. Exim crashed if a remote delivery caused a very long error message to be
4810 recorded - for instance if somebody sent an entire SpamAssassin report back
4811 as a large number of 550 error lines. This bug was coincidentally fixed by
4812 increasing the size of one of Exim's internal buffers (big_buffer) that
4813 happened as part of the Exiscan merge. However, to be on the safe side, I
4814 have made the code more robust (and fixed the comments that describe what
4817 53. Now that there can be additional text after "Completed" in log lines (if
4818 the queue_time_overall log selector is set), a one-byte patch to exigrep
4819 was needed to allow it to recognize "Completed" as not the last thing in
4822 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4823 patch that reportedly fixes this has been added. I am not expert enough to
4824 create a test for it. This is what the patch creator wrote:
4826 "I found a little strange behaviour of ldap code when working with
4827 Windows 2003 AD Domain, where users was placed in more than one
4828 Organization Units. When I tried to give exim partial DN, the exit code
4829 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4830 But simultaneously result of request was absolutely normal ldap result,
4831 so I produce this patch..."
4833 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4834 REFERENCE, so I have modified the code to exclude the patch when that macro
4837 55. Some experimental protocols are using DNS PTR records for new purposes. The
4838 keys for these records are domain names, not reversed IP addresses. The
4839 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4840 leaves it alone. Component reversal etc. now happens only for IP addresses.
4843 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4845 57. Double the size of the debug message buffer (to 2048) so that more of very
4846 long debug lines gets shown.
4848 58. The exicyclog utility now does better if the number of log files to keep
4849 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4851 59. Two changes related to the smtp_active_hostname option:
4853 (1) $smtp_active_hostname is now available as a variable.
4854 (2) The default for smtp_banner uses $smtp_active_hostname instead
4855 of $primary_hostname.
4857 60. The host_aton() function is supposed to be passed a string that is known
4858 to be a valid IP address. However, in the case of IPv6 addresses, it was
4859 not checking this. This is a hostage to fortune. Exim now panics and dies
4860 if the condition is not met. A case was found where this could be provoked
4861 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4862 components; fortuitously, this particular loophole had already been fixed
4863 by change 4.50/55 above.
4865 If there are any other similar loopholes, the new check in host_aton()
4866 itself should stop them being exploited. The report I received stated that
4867 data on the command line could provoke the exploit when Exim was running as
4868 exim, but did not say which command line option was involved. All I could
4869 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4870 running as the user.
4873 61. There was a buffer overflow vulnerability in the SPA authentication code
4874 (which came originally from the Samba project). I have added a test to the
4875 spa_base64_to_bits() function which I hope fixes it.
4878 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4879 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4881 63. The daemon start-up calls getloadavg() while still root for those OS that
4882 need the first call to be done as root, but it missed one case: when
4883 deliver_queue_load_max is set with deliver_drop_privilege. This is
4884 necessary for the benefit of the queue runner, because there is no re-exec
4885 when deliver_drop_privilege is set.
4887 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4888 This has been fixed.
4890 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4891 was in use, was not putting the data itself into the right store pool;
4892 consequently, it could be overwritten for a subsequent message in the same
4893 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4896 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4898 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4899 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4900 the function were treating the return as a boolean value, which happened to
4901 work because 0=false and not-0=true, but is not correct code.
4903 68. The host_aton() function was not handling scoped IPv6 addresses (those
4904 with, for example, "%eth0" on the end) correctly.
4906 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4907 negated items (that is, ~something) in unsigned ints. Some compilers
4908 apparently mutter when there is no cast.
4910 70. If an address verification called from an ACL failed, and did not produce a
4911 user-specific message (i.e. there was only a "system" message), nothing was
4912 put in $acl_verify_message. In this situation, it now puts the system
4915 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4916 session; change 4.31/43 added the unwanted input to the log line - except
4917 that it did not do this in the start of session case. It now does.
4919 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4920 This is wrong and can cause the other end to generate a synchronization
4921 error if it is another Exim or anything else that does the synchronization
4922 check. A QUIT command is no longer sent after a timeout.
4924 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4925 during host lookups.
4927 74. The defer_ok option of callout verification was not working if it was used
4928 when verifying addresses in header lines, that is, for this case:
4930 verify = header_sender/callout=defer_ok
4932 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4933 those file descriptors could be used for SMTP connections. If anything
4934 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4935 could be sent to the SMTP client, causing chaos. The daemon now opens
4936 stdin, stdout, and stderr to /dev/null when it puts itself into the
4939 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4940 log by default. The user can override this with suitable Perl magic.
4942 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4943 the log message when discard triggers, was not working for the DATA ACL or
4944 for the non-SMTP ACL.
4946 78. Error message wording change in sieve.c.
4948 79. If smtp_accept_max_per_host was set, the number of connections could be
4949 restricted to fewer than expected, because the daemon was trying to set up
4950 a new connection before checking whether the processes handling previous
4951 connections had finished. The check for completed processes is now done
4952 earlier. On busy systems, this bug wouldn't be noticed because something
4953 else would have woken the daemon, and it would have reaped the completed
4956 80. If a message was submitted locally by a user whose login name contained one
4957 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4958 It caused a spool format error. I have fixed the spool reading code. A
4959 related problem was that the "from" clause in the Received: line became
4960 illegal because of the space(s). It is now covered by ${quote_local_part.
4962 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4965 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4967 83. Patch from Sieve maintainer to fix unterminated string problem in
4968 "vacation" handling.
4970 84. Some minor changes to the Linux configuration files to help with other
4971 OS variants using glibc.
4973 85. One more patch for Sieve to update vacation handling to latest spec.
4976 ----------------------------------------------------
4977 See the note above about the 4.44 and 4.50 releases.
4978 ----------------------------------------------------
4984 1. Change 4.43/35 introduced a bug that caused file counts to be
4985 incorrectly computed when quota_filecount was set in an appendfile
4988 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4989 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4992 3. Give more explanation in the error message when the command for a transport
4993 filter fails to execute.
4995 4. There are several places where Exim runs a non-Exim command in a
4996 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4997 was being done only for the command run by the queryprogram router. It is
4998 now done for all such subprocesses. The other cases are: ${run, transport
4999 filters, and the commands run by the lmtp and pipe transports.
5001 5. Some older OS have a limit of 256 on the maximum number of file
5002 descriptors. Exim was using setrlimit() to set 1000 as a large value
5003 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5004 systems. I've change it so that if it can't get 1000, it tries for 256.
5006 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5007 was an oversight, and furthermore, ever since the addition of extra
5008 controls (e.g. 4.43/32), the checks on when to allow different forms of
5009 "control" were broken. There should now be diagnostics for all cases when a
5010 control that does not make sense is encountered.
5012 7. $recipients is now available in the predata ACL (oversight).
5014 8. Tidy the search cache before the fork to do a delivery from a message
5015 received from the command line. Otherwise the child will trigger a lookup
5016 failure and thereby defer the delivery if it tries to use (for example) a
5017 cached ldap connection that the parent has called unbind on.
5019 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5020 of $address_data from the recipient verification was clobbered by the
5021 sender verification.
5023 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5024 was its contents. (It was OK if the option was not defined at all.)
5026 11. A "Completed" log line is now written for messages that are removed from
5027 the spool by the -Mrm option.
5029 12. $host_address is now set to the target address during the checking of
5030 ignore_target_hosts.
5032 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5033 being passed; this would have caused $sender_host_name to have been used if
5034 matching the list had actually called for a host name (not very likely,
5035 since this list is usually IP addresses). A host name is now passed as
5038 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5039 code that specifies a non-restarting handler (typically sigaction() in
5040 modern systems) in an attempt to fix a rare and obscure crash bug.
5042 15. Narrowed the window for a race in the daemon that could cause it to ignore
5043 SIGCHLD signals. This is not a major problem, because they are used only to
5044 wake it up if nothing else does.
5046 16. A malformed maildirsize file could cause Exim to calculate negative values
5047 for the mailbox size or file count. Odd effects could occur as a result.
5048 The maildirsize information is now recalculated if the size or filecount
5051 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5052 support for a long time. Removed HAVE_SYS_VFS_H.
5054 18. Updated exipick to current release from John Jetmore.
5056 19. Allow an empty sender to be matched against a lookup in an address list.
5057 Previously the only cases considered were a regular expression, or an
5060 20. Exim went into a mad DNS lookup loop when doing a callout where the
5061 host was specified on the transport, if the DNS lookup yielded more than
5064 21. The RFC2047 encoding function was originally intended for short strings
5065 such as real names; it was not keeping to the 75-character limit for
5066 encoded words that the RFC imposes. It now respects the limit, and
5067 generates multiple encoded words if necessary. To be on the safe side, I
5068 have increased the buffer size for the ${rfc2047: expansion operator from
5071 22. Failure to deliver a bounce message always caused it to be frozen, even if
5072 there was an errors_to setting on the router. The errors_to setting is now
5075 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5076 canonical form (fully expanded) before being placed in
5077 $sender_host_address.
5079 24. Updated eximstats to version 1.33
5081 25. Include certificate and key file names in error message when GnuTLS fails
5082 to set them up, because the GnuTLS error message doesn't include the name
5083 of the failing file when there is a problem reading it.
5085 26. Expand error message when OpenSSL has problems setting up cert/key files.
5088 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5089 (this can affect the format of dates).
5091 28. exim_tidydb, when checking for the continued existence of a message for
5092 which it has found a message-specific retry record, was not finding
5093 messages that were in split spool directories. Consequently, it was
5094 deleting retry records that should have stayed in existence.
5096 29. eximstats updated to version 1.35
5097 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5098 1.35 - bugfix such that pie charts by volume are generated correctly
5100 30. The SPA authentication driver was not abandoning authentication and moving
5101 on to the next authenticator when an expansion was forced to fail,
5102 contradicting the general specification for all authenticators. Instead it
5103 was generating a temporary error. It now behaves as specified.
5105 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5106 (the order specifies the preference for clients). The order is now AES256,
5107 AES128, 3DES, ARCFOUR128.
5109 31. Small patch to Sieve code - explicitly set From: when generating an
5112 32. Exim crashed if a remote delivery caused a very long error message to be
5113 recorded - for instance if somebody sent an entire SpamAssassin report back
5114 as a large number of 550 error lines. This bug was coincidentally fixed by
5115 increasing the size of one of Exim's internal buffers (big_buffer) that
5116 happened as part of the Exiscan merge. However, to be on the safe side, I
5117 have made the code more robust (and fixed the comments that describe what
5120 33. Some experimental protocols are using DNS PTR records for new purposes. The
5121 keys for these records are domain names, not reversed IP addresses. The
5122 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5123 leaves it alone. Component reversal etc. now happens only for IP addresses.
5126 34. The host_aton() function is supposed to be passed a string that is known
5127 to be a valid IP address. However, in the case of IPv6 addresses, it was
5128 not checking this. This is a hostage to fortune. Exim now panics and dies
5129 if the condition is not met. A case was found where this could be provoked
5130 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5131 components; fortuitously, this particular loophole had already been fixed
5132 by change 4.50/55 or 4.44/33 above.
5134 If there are any other similar loopholes, the new check in host_aton()
5135 itself should stop them being exploited. The report I received stated that
5136 data on the command line could provoke the exploit when Exim was running as
5137 exim, but did not say which command line option was involved. All I could
5138 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5139 running as the user.
5142 35. There was a buffer overflow vulnerability in the SPA authentication code
5143 (which came originally from the Samba project). I have added a test to the
5144 spa_base64_to_bits() function which I hope fixes it.
5147 36. The daemon start-up calls getloadavg() while still root for those OS that
5148 need the first call to be done as root, but it missed one case: when
5149 deliver_queue_load_max is set with deliver_drop_privilege. This is
5150 necessary for the benefit of the queue runner, because there is no re-exec
5151 when deliver_drop_privilege is set.
5153 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5154 was in use, was not putting the data itself into the right store pool;
5155 consequently, it could be overwritten for a subsequent message in the same
5156 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5159 38. Sometimes the final signoff response after QUIT could fail to get
5160 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5161 < 0 before doing a fflush(). This bug looks as though it goes back to the
5162 introduction of TLS in release 3.20, but "sometimes" must have been rare
5163 because the tests only now provoked it.
5169 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5170 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5171 Now it returns an integer. A number of calls were still expecting a T/F
5172 return. Fortuitously, in all cases, the tests worked in OK situations,
5173 which is the norm. However, things would have gone wrong on any write
5174 failures on the smtp file descriptor. This function is used when sending
5175 messages over SMTP and also when doing verify callouts.
5177 2. When Exim is called to do synchronous delivery of a locally submitted
5178 message (the -odf or -odi options), it no longer closes stderr before doing
5181 3. Implemented the mua_wrapper option.
5183 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5185 5. Implemented the functions header_remove(), header_testname(),
5186 header_add_at_position(), and receive_remove_recipient(), and exported them
5189 6. If an ACL "warn" statement specified the addition of headers, Exim already
5190 inserted X-ACL-Warn: at the start if there was no header name. However, it
5191 was not making this test for the second and subsequent header lines if
5192 there were newlines in the string. This meant that an invalid header could
5193 be inserted if Exim was badly configured.
5195 7. Allow an ACL "warn" statement to add header lines at the start or after all
5196 the Received: headers, as well as at the end.
5198 8. Added the rcpt_4xx retry error code.
5200 9. Added postmaster_mailfrom=xxx to callout verification option.
5202 10. Added mailfrom=xxxx to the callout verification option, for verify=
5205 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5206 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5208 12. Inserted some casts to stop certain compilers warning when using pointer
5209 differences as field lengths or precisions in printf-type calls (mostly
5210 affecting debugging statements).
5212 13. Added optional readline() support for -be (dynamically loaded).
5214 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5215 same clock tick as a message's arrival, so that its received time was the
5216 same as the "first fail" time on the retry record, and that message
5217 remained on the queue past the ultimate address timeout, every queue runner
5218 would try a delivery (because it was past the ultimate address timeout) but
5219 after another failure, the ultimate address timeout, which should have then
5220 bounced the address, did not kick in. This was a "< instead of <=" error;
5221 in most cases the first failure would have been in the next clock tick
5222 after the received time, and all would be well.
5224 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5225 being recognized when the domain list was tested by the match_domain
5226 condition in an expansion string.
5228 16. Added the ${str2b64: operator.
5230 17. Exim was always calling setrlimit() to set a large limit for the number of
5231 processes, without checking whether the existing limit was already
5232 adequate. (It did check for the limit on file descriptors.) Furthermore,
5233 errors from getrlimit() and setrlimit() were being ignored. Now they are
5234 logged to the main and panic logs, but Exim does carry on, to try to do its
5235 job under whatever limits there are.
5237 18. Imported PCRE 5.0.
5239 19. Trivial typo in log message " temporarily refused connection" (the leading
5242 20. If the log selector return_path_on_delivery was set and an address was
5243 redirected to /dev/null, the delivery process crashed because it assumed
5244 that a return path would always be set for a "successful" delivery. In this
5245 case, the whole delivery is bypassed as an optimization, and therefore no
5248 21. Internal re-arrangement: the function for sending a challenge and reading
5249 a response while authentication was assuming a zero-terminated challenge
5250 string. It's now changed to take a pointer and a length, to allow for
5251 binary data in such strings.
5253 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5255 23. Exim was not respecting finduser_retries when seeking the login of the
5256 uid under which it was called; it was always trying 10 times. (The default
5257 setting of finduser_retries is zero.) Also, it was sleeping after the final
5258 failure, which is pointless.
5260 24. Implemented tls_on_connect_ports.
5262 25. Implemented acl_smtp_predata.
5264 26. If the domain in control=submission is set empty, Exim assumes that the
5265 authenticated id is a complete email address when it generates From: or
5266 Sender: header lines.
5268 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5269 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5270 chown and chgrp in /bin and hostname in /usr/bin.
5272 28. Exim was keeping the "process log" file open after each use, just as it
5273 does for the main log. This opens the possibility of it remaining open for
5274 long periods when the USR1 signal hits a daemon. Occasional processlog
5275 errors were reported, that could have been caused by this. Anyway, it seems
5276 much more sensible not to leave this file open at all, so that is what now
5279 29. The long-running daemon process does not normally write to the log once it
5280 has entered its main loop, and it closes the log before doing so. This is
5281 so that log files can straightforwardly be renamed and moved. However,
5282 there are a couple of unusual error situations where the daemon does write
5283 log entries, and I had neglected to close the log afterwards.
5285 30. The text of an SMTP error response that was received during a remote
5286 delivery was being truncated at 512 bytes. This is too short for some of
5287 the long messages that one sometimes sees. I've increased the limit to
5290 31. It is now possible to make retry rules that apply only when a message has a
5291 specific sender, in particular, an empty sender.
5293 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5294 it possible to be selective about when SMTP synchronization is enforced.
5296 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5298 32. Implemented hosts_connection_nolog.
5300 33. Added an ACL for QUIT.
5302 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5305 35. Added mailbox_size and mailbox_filecount to appendfile.
5307 36. Added control = no_multiline_responses to ACLs.
5309 37. There was a bug in the logic of the code that waits for the clock to tick
5310 in the case where the clock went backwards by a substantial amount such
5311 that the microsecond fraction of "now" was more than the microsecond
5312 fraction of "then" (but the whole seconds number was less).
5314 38. Added support for the libradius Radius client library this is found on
5315 FreeBSD (previously only the radiusclient library was supported).
5321 1. When certain lookups returned multiple values in the form name=value, the
5322 quoting of the values was not always being done properly. Specifically:
5323 (a) If the value started with a double quote, but contained no whitespace,
5325 (b) If the value contained whitespace other than a space character (i.e.
5326 tabs or newlines or carriage returns) it was not quoted.
5327 This fix has been applied to the mysql and pgsql lookups by writing a
5328 separate quoting function and calling it from the lookup code. The fix
5329 should probably also be applied to nisplus, ibase and oracle lookups, but
5330 since I cannot test any of those, I have not disturbed their existing code.
5332 2. A hit in the callout cache for a specific address caused a log line with no
5333 reason for rejecting RCPT. Now it says "Previous (cached) callout
5334 verification failure".
5336 3. There was an off-by-one bug in the queryprogram router. An over-long
5337 return line was truncated at 256 instead of 255 characters, thereby
5338 overflowing its buffer with the terminating zero. As well as fixing this, I
5339 have increased the buffer size to 1024 (and made a note to document this).
5341 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5342 when Exim is waiting for an SMTP response from a remote server, Exim
5343 restarts its select() call on the socket, thereby resetting its timeout.
5344 This is not a problem when such interrupts are rare. Somebody set up a cron
5345 job to run exiwhat every 2 minutes, which is less than the normal select()
5346 timeout (5 or 10 minutes). This meant that the select() timeout never
5347 kicked in because it was always reset. I have fixed this by comparing the
5348 time when an interrupt arrives with the time at the start of the first call
5349 to select(). If more time than the timeout has elapsed, the interrupt is
5350 treated as a timeout.
5352 5. Some internal re-factoring in preparation for the addition of Sieve
5353 extensions (by MH). In particular, the "personal" test is moved to a
5354 separate function, and given an option for scanning Cc: and Bcc: (which is
5355 not set for Exim filters).
5357 6. When Exim created an email address using the login of the caller as the
5358 local part (e.g. when creating a From: or Sender: header line), it was not
5359 quoting the local part when it contained special characters such as @.
5361 7. Installed new OpenBSD configuration files.
5363 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5364 try to make them clearer.
5366 9. Callout options, other than the timeout value, were being ignored when
5367 verifying sender addresses in header lines. For example, when using
5369 verify = header_sender/callout=no_cache
5371 the cache was (incorrectly) being used.
5373 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5374 only the Cygwin environment.
5376 11. When return_path_on_delivery was set as a log selector, if different remote
5377 addresses in the same message used different return paths and parallel
5378 remote delivery occurred, the wrong values would sometimes be logged.
5379 (Whenever a remote delivery process finished, the return path value from
5380 the most recently started remote delivery process was logged.)
5382 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5383 lines when AUTH and/or TLS are in use. This is the "received protocol"
5384 field. Exim used to use "asmtp" for authenticated SMTP, without any
5385 indication (in the protocol name) for TLS use. Now it follows the RFC and
5386 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5387 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5388 names appear in log lines as well as in Received: header lines.
5390 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5391 extensions, and comparison tests, and to fix some bugs.
5393 14. Changes to the "personal" filter test:
5395 (1) The test was buggy in that it was just doing the equivalent of
5396 "contains" tests on header lines. For example, if a user's address was
5397 anne@some.where, the "personal" test would incorrectly be true for
5399 To: susanne@some.where
5401 This test is now done by extracting each address from the header in turn,
5402 and checking the entire address. Other tests that are part of "personal"
5403 are now done using regular expressions (for example, to check local parts
5404 of addresses in From: header lines).
5406 (2) The list of non-personal local parts in From: addresses has been
5407 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5408 taken from the Sieve specification recommendations.
5410 (3) If the message contains any header line starting with "List-" it is
5411 treated as non-personal.
5413 (4) The test for "circular" in the Subject: header line has been removed
5414 because it now seems ill-conceived.
5416 15. Minor typos in src/EDITME comments corrected.
5418 16. Installed latest exipick from John Jetmore.
5420 17. If headers_add on a router specified a text string that was too long for
5421 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5422 of string_sprintf() is now avoided.
5424 18. $message_body_size was not set (it was always zero) when running the DATA
5425 ACL and the local_scan() function.
5427 19. For the "mail" command in an Exim filter, no default was being set for
5428 the once_repeat time, causing a random time value to be used if "once" was
5429 specified. (If the value happened to be <= 0, no repeat happened.) The
5430 default is now 0s, meaning "never repeat". The "vacation" command was OK
5431 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5432 (I found it when inspecting the code).
5434 20. There is now an overall timeout for performing a callout verification. It
5435 defaults to 4 times the callout timeout, which applies to individual SMTP
5436 commands during the callout. The overall timeout applies when there is more
5437 than one host that can be tried. The timeout is checked before trying the
5438 next host. This prevents very long delays if there are a large number of
5439 hosts and all are timing out (e.g. when the network connections are timing
5440 out). The value of the overall timeout can be changed by specifying an
5441 additional sub-option for "callout", called "maxwait". For example:
5443 verify = sender/callout=5s,maxwait=20s
5445 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5446 to the end before writing, but this should make it even safer).
5448 22. Exim was forgetting that it had advertised PIPELINING for the second and
5449 subsequent messages on an SMTP connection. It was also not resetting its
5450 memory on STARTTLS and an internal HELO.
5452 23. When Exim logs an SMTP synchronization error within a session, it now
5453 records whether PIPELINING has been advertised or not.
5455 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5456 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5457 rather than long int.
5459 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5461 26. Added the never_mail option to autoreply.
5467 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5468 crash if the getsockname() call failed; this can happen if a connection is
5469 closed very soon after it is established. The problem was simply in the
5470 order in which certain operations were done, causing Exim to try to write
5471 to the SMTP stream before it had set up the file descriptor. The bug has
5472 been fixed by making things happen in the correct order.
5478 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5479 before the connection was closed, thus losing the rejection response.
5481 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5482 some early Solaris releases, but causes trouble in current releases where
5483 socklen_t is defined.
5485 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5488 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5491 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5492 that releasing the top of it at the end releases what was used for sub-
5493 expansions (unless the block got too big). However, discard this block if
5494 the first thing is a variable or header, so that we can use its block when
5495 it is dynamic (useful for very large $message_headers, for example).
5497 6. Lookups now cache *every* query, not just the most recent. A new, separate
5498 store pool is used for this. It can be recovered when all lookup caches are
5499 flushed. Lookups now release memory at the end of their result strings.
5500 This has involved some general refactoring of the lookup sources.
5502 7. Some code has been added to the store_xxx() functions to reduce the amount
5503 of flapping under certain conditions.
5505 8. log_incoming_interface used to affect only the <= reception log lines. Now
5506 it causes the local interface and port to be added to several more SMTP log
5507 lines, for example "SMTP connection from", and rejection lines.
5509 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5511 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5513 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5514 could be overwritten at the end of the current message (or the start of a
5515 new message if it was set in a HELO ACL). The value is now preserved for
5516 the duration of the SMTP connection.
5518 12. If a transport had a headers_rewrite setting, and a matching header line
5519 contained an unqualified address, that address was qualified, even if it
5520 did not match any rewriting rules. The underlying bug was that the values
5521 of the flags that permit the existence of unqualified sender and recipient
5522 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5523 non-local messages, and by -bnq for local messages) were not being
5524 preserved with the message after it was received.
5526 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5527 "next input=" as part of the text comprising the host identity instead of
5528 the correct text. The code was using the same buffer for two different
5529 strings. However, depending on which order the printing function evaluated
5530 its arguments, the bug did not always show up. Under Linux, for example, my
5531 test suite worked just fine.
5533 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5534 doesn't work with some older versions of Perl. It has been changed to "my",
5535 which in any case is probably the better facility to use.
5537 15. A really picky compiler found some instances of statements for creating
5538 error messages that either had too many or two few arguments for the format
5541 16. The size of the buffer for calls to the DNS resolver has been increased
5542 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5543 for addresses that have a lot of PTR records. This alleviates a problem; it
5544 does not fully solve it.
5546 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5547 buffer now truncates the list and logs the incident, which is the same
5548 action as happens when Exim is looking up a host name and its aliases.
5549 Previously in this situation something unpredictable would happen;
5550 sometimes it was "internal error: store_reset failed".
5552 18. If a server dropped the connection unexpectedly when an Exim client was
5553 using GnuTLS and trying to read a response, the client delivery process
5554 crashed while trying to generate an error log message.
5556 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5557 string, for example:
5559 warn message = H1: something\nH2: something
5561 the text was added as a single header line from Exim's point of view
5562 though it ended up OK in the delivered message. However, searching for the
5563 second and subsequent header lines using $h_h2: did not work. This has been
5564 fixed. Similarly, if a system filter added multiple headers in this way,
5565 the routers could not see them.
5567 20. Expanded the error message when iplsearch is called with an invalid key to
5568 suggest using net-iplsearch in a host list.
5570 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5571 ACLs are no longer actually imposed (and a message to that effect is
5574 22. If a "gecos" field in a passwd entry contained escaped characters, in
5575 particular, if it contained a \" sequence, Exim got it wrong when building
5576 a From: or a Sender: header from that name. A second bug also caused
5577 incorrect handling when an unquoted " was present following a character
5578 that needed quoting.
5580 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5581 was not being matched caselessly.
5583 24. Arranged for all hyphens in the exim.8 source to be escaped with
5586 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5587 properly. Recipient callout cache records were still being keyed to include
5588 the sender, even when use_sender was set false. This led to far more
5589 callouts that were necessary. The sender is no longer included in the key
5590 when use_sender is false.
5592 26. Added "control = submission" modifier to ACLs.
5594 27. Added the ${base62d: operator to decode base 62 numbers.
5596 28. dnsdb lookups can now access SRV records.
5598 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5599 the configuration file.
5601 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5602 (-v) mode. This makes the output for a verbose queue run more intelligible.
5604 31. Added a use_postmaster feature to recipient callouts.
5606 32. Added the $body_zerocount variable, containing the number of binary zero
5607 bytes in the message body.
5609 33. The time of last modification of the "new" subdirectory is now used as the
5610 "mailbox time last read" when there is a quota error for a maildir
5613 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5615 35. Added +ignore_unknown as a special item in host lists.
5617 36. Code for decoding IPv6 addresses in host lists is now included, even if
5618 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5619 address was recognized as an IP address, but was then not correctly decoded
5620 into binary, causing unexpected and incorrect effects when compared with
5627 1. Very minor rewording of debugging text in manualroute to say "list of
5628 hosts" instead of "hostlist".
5630 2. If verify=header_syntax was set, and a header line with an unqualified
5631 address (no domain) and a large number of spaces between the end of the
5632 name and the colon was received, the reception process suffered a buffer
5633 overflow, and (when I tested it) crashed. This was caused by some obsolete
5634 code that should have been removed. The fix is to remove it!
5636 3. When running in the test harness, delay a bit after writing a bounce
5637 message to get a bit more predictability in the log output.
5639 4. Added a call to search_tidyup() just before forking a reception process. In
5640 theory, someone could use a lookup in the expansion of smtp_accept_max_
5641 per_host which, without the tidyup, could leave open a database connection.
5643 5. Added the variables $recipient_data and $sender_data which get set from a
5644 lookup success in an ACL "recipients" or "senders" condition, or a router
5645 "senders" option, similar to $domain_data and $local_part_data.
5647 6. Moved the writing of debug_print from before to after the "senders" test
5650 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5651 problems for message scanning, either using a data ACL, or using
5652 local_scan() because the Received: header was not generated till after they
5653 were called (in order to set the time as the time of reception completion).
5654 I have revised the way this works. The header is now generated after the
5655 body is received, but before the ACL or local_scan() are called. After they
5656 are run, the timestamp in the header is updated.
5662 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5663 before starting a queue runner without re-exec. This happened only when
5664 deliver_drop_privilege was set or when the Exim user was set to root. The
5665 effect of the bug was that timeouts during subsequent deliveries caused
5666 crashes instead of being properly handled. The handler is now left at its
5667 default (and expected) setting.
5669 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5670 message, again when deliver_drop_privilege is set or Exim is run as root.
5671 The bug described in (1) was not present in this case, but the tidying up
5672 of the other signals was missing. I have made the two cases consistent.
5674 3. The ignore_target_hosts setting on a manualroute router was being ignored
5675 for hosts that were looked up using the /MX notation.
5677 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5680 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5681 operated on the sender address. After changing the $sender_address to <>
5682 for the sender address verify, Exim was re-instated it as the original
5683 (before rewriting) address, but remembering that it had rewritten it, so it
5684 wasn't rewriting it again. This bug also had the effect of breaking the
5685 sender address verification caching when the sender address was rewritten.
5687 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5688 This has been changed so that if the ip literal address matches
5689 ignore_target_hosts, the router declines.
5691 7. Added expansion conditions match_domain, match_address, and match_local_
5692 part (NOT match_host).
5694 8. The placeholder for the Received: header didn't have a length field set.
5696 9. Added code to Exim itself and to exim_lock to test for a specific race
5697 condition that could lead to file corruption when using MBX delivery. The
5698 issue is with the lockfile that is created in /tmp. If this file is removed
5699 after a process has opened it but before that process has acquired a lock,
5700 there is the potential for a second process to recreate the file and also
5701 acquire a lock. This could lead to two Exim processes writing to the file
5702 at the same time. The added code performs the same test as UW imapd; it
5703 checks after acquiring the lock that its file descriptor still refers to
5704 the same named file.
5706 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5707 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5710 11. Added the smtp_active_hostname option. If used, this will typically be made
5711 to depend on the incoming interface address. Because $interface_address is
5712 not set up until the daemon has forked a reception process, error responses
5713 that can happen earlier (such as "too many connections") no longer contain
5716 12. If an expansion in a condition on a "warn" statement fails because a lookup
5717 defers, the "warn" statement is abandoned, and the next ACL statement is
5718 processed. Previously this caused the whole ACL to be aborted.
5720 13. Added the iplsearch lookup type.
5722 14. Added ident_timeout as a log selector.
5724 15. Added tls_certificate_verified as a log selector.
5726 16. Added a global option tls_require_ciphers (compare the smtp transport
5727 option of the same name). This controls incoming TLS connections.
5729 17. I finally figured out how to make tls_require_ciphers do a similar thing
5730 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5731 before starting the TLS session.
5733 18. Tabs are now shown as \t in -bP output.
5735 19. If the log selector return_path_on_delivery was set, Exim crashed when
5736 bouncing a message because it had too many Received: header lines.
5738 20. If two routers both had headers_remove settings, and the first one included
5739 a superfluous trailing colon, the final name in the first list and the
5740 first name in the second list were incorrectly joined into one item (with a
5741 colon in the middle).
5747 1. Added -C and -D options to the exinext utility, mainly to make it easier
5748 to include in the automated testing, but these could be helpful when
5749 multiple configurations are in use.
5751 2. The exinext utility was not formatting the output nicely when there was
5752 an alternate port involved in the retry record key, nor when there was a
5753 message id as well (for retries that were specific to a specific message
5754 and a specific host). It was also confused by IPv6 addresses, because of
5755 the additional colons they contain. I have fixed the IPv4 problem, and
5756 patched it up to do a reasonable job for IPv6.
5758 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5759 delivery, the log line now contains "pipelined" if PIPELINING was used.
5761 4. An SMTP transport process used to panic and die if the bind() call to set
5762 an explicit outgoing interface failed. This has been changed; it is now
5763 treated in the same way as a connect() failure.
5765 5. A reference to $sender_host_name in the part of a conditional expansion
5766 that was being skipped was still causing a DNS lookup. This no longer
5769 6. The def: expansion condition was not recognizing references to header lines
5770 that used bh_ and bheader_.
5772 7. Added the _cache feature to named lists.
5774 8. The code for checking quota_filecount in the appendfile transport was
5775 allowing one more file than it should have been.
5777 9. For compatibility with Sendmail, the command line option
5785 and sets the incoming protocol and host name (for trusted callers). The
5786 host name and its colon can be omitted when only the protocol is to be set.
5787 Note the Exim already has two private options, -pd and -ps, that refer to
5788 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5789 "s", but I don't think that's a major issue.
5791 10. A number of refactoring changes to the code, none of which should affect
5794 (a) The number of logging options was getting close to filling up the
5795 32-bit word that was used as a bit map. I have split them into two classes:
5796 those that are passed in the argument to log_write(), and those that are
5797 only ever tested independently outside of that function. These are now in
5798 separate 32-bit words, so there is plenty of room for expansion again.
5799 There is no change in the user interface or the logging behaviour.
5801 (b) When building, for example, log lines, the code previously used a
5802 macro that called string_cat() twice, in order to add two strings. This is
5803 not really sufficiently general. Furthermore, there was one instance where
5804 it was actually wrong because one of the argument was used twice, and in
5805 one call a function was used. (As it happened, calling the function twice
5806 did not affect the overall behaviour.) The macro has been replaced by a
5807 function that can join an arbitrary number of extra strings onto a growing
5810 (c) The code for expansion conditions now uses a table and a binary chop
5811 instead of a serial search (which was left over from when there were very
5812 few conditions). Also, it now recognizes conditions like "pam" even when
5813 the relevant support is not compiled in: a suitably worded error message is
5814 given if an attempt is made to use such a condition.
5816 11. Added ${time_interval:xxxxx}.
5818 12. A bug was causing one of the ddress fields not to be passed back correctly
5819 from remote delivery subprocesses. The field in question was not being
5820 subsequently used, so this caused to problems in practice.
5822 13. Added new log selectors queue_time and deliver_time.
5824 14. Might have fixed a bug in maildirsizefile handling that threw up
5825 "unexpected character" debug warnings, and recalculated the data
5826 unnecessarily. In any case, I expanded the warning message to give more
5829 15. Added the message "Restricted characters in address" to the statements in
5830 the default ACL that block characters like @ and % in local parts.
5832 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5833 Three changes have been made:
5835 (a) There was a serious bug; a negative response to MAIL caused the whole
5836 recipient domain to be cached as invalid, thereby blocking all messages
5837 to all local parts at the same domain, from all senders. This bug has
5838 been fixed. The domain is no longer cached after a negative response to
5839 MAIL if the sender used is not empty.
5841 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5844 (c) A new callout option, "use_sender" has been added for people who want
5845 the modified behaviour.
5851 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5854 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5855 indeed breaks things for older releases.
5857 3. Added additional logging to the case where there is a problem reading data
5858 from a filter that is running in a subprocess using a pipe, in order to
5859 try to track down a specific problem.
5861 4. Testing facility fudge: when running in the test harness and attempting
5862 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5863 getting "No route to host". Convert this to a timeout.
5865 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5868 6. Some OS don't have socklen_t but use size_t instead. This affects the
5869 fifth argument of getsockopt() amongst other things. This is now
5870 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5871 can be set for individual OS. I have set it for SunOS5, OSF1, and
5872 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5873 some earlier ones do not.
5875 7. Change 4.30/15 was not doing the test caselessly.
5877 8. The standard form for an IPv6 address literal was being rejected by address
5878 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5879 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5880 this, as well as the form without the "IPv6" on the front (but only when
5881 address literals are enabled, of course).
5883 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5885 10. Exim crashed if a message with an empty sender address specified by -f
5886 encountered a router with an errors_to setting. This could be provoked only
5887 by a command such as
5891 where an empty string was supplied; "<>" did not hit this bug.
5893 11. Installed PCRE release 4.5.
5895 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5896 remained set. It is now erased.
5898 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5899 times from message ids (which are base 36 rather than the normal 62).
5901 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5902 were being counted as actual protocol errors, and logged if the log
5903 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5904 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5905 and DATA following a set of rejected RCPTs do not count as protocol errors.
5906 In other words, Exim assumes they were pipelined, though this may not
5907 actually be the case. Of course, in all cases the client gets an
5908 appropriate error code.
5910 15. If a lookup fails in an ACL condition, a message about the failure may
5911 be available; it is used if testing the ACL cannot continue, because most
5912 such messages specify what the cause of the deferral is. However, some
5913 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5914 that caused an old message to be retained and used if a later statement
5915 caused a defer, replacing the real cause of the deferral.
5917 16. If an IP address had so many PTR records that the DNS lookup buffer
5918 was not large enough to hold them, Exim could crash while trying to process
5919 the truncated data. It now detects and logs this case.
5921 17. Further to 4.21/58, another change has been made: if (and only if) the
5922 first line of a message (the first header line) ends with CRLF, a bare LF
5923 in a subsequent header line has a space inserted after it, so as not to
5924 terminate the header.
5926 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5927 unnecessarily, used atoi() instead of strtol(), and didn't check the
5928 termination when getting file sizes from file names by regex.
5930 19. Completely re-implemented the support for maildirsize files, in the light
5931 of a number of problems with the previous contributed implementation
5932 (4.30/29). In particular:
5934 . If the quota is zero, the maildirsize file is maintained, but no quota is
5937 . If the maildir directory does not exist, it is created before any attempt
5938 to write a maildirsize file.
5940 . The quota value in the file is just a cache; if the quota is changed in
5941 the transport, the new value overrides.
5943 . A regular expression is available for excluding directories from the
5946 20. The autoreply transport checks the characters in options that define the
5947 message's headers; it allows continued headers, but it was checking with
5948 isspace() after an embedded newline instead of explicitly looking for a
5951 21. If all the "regular" hosts to which an address was routed had passed their
5952 expiry times, and had not reached their retry times, the address was
5953 bounced, even if fallback hosts were defined. Now Exim should go on to try
5956 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5957 equivalent code in the SMTP transport. Some hosts send humungous responses
5958 to HELO/EHLO, more than 1024 it seems.
5960 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5961 gives compiler warnings in some environments. I've now done it "properly",
5964 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5965 (because of problems with the built-in one) was declared to return uschar *
5966 instead of char *, causing compiler failure.
5968 25. Fixed a file descriptor leak when processing alias/forward files.
5970 26. Fixed a minor format string issue in dbfn.c.
5972 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5974 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5975 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5978 29. When a sender address is verified, it is cached, to save repeating the test
5979 when there is more than one recipient in a message. However, when the
5980 verification involves a callout, it is possible for different callout
5981 options to be set for different recipients. It is too complicated to keep
5982 track of this in the cache, so now Exim always runs a verification when a
5983 callout is required, relying on the callout cache for the optimization.
5984 The overhead is duplication of the address routing, but this should not be
5987 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5988 to be verified with callout=postmaster, and the main callout worked but the
5989 postmaster check failed, the verification correctly failed. However, if a
5990 subsequent RCPT command asked for sender verification *without* the
5991 postmaster check, incorrect caching caused this verification also to fail,
5994 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5995 it was not caching the DNS options (qualify_single, search_parents) that
5996 were used when the lookup failed. A subsequent lookup with different
5997 options therefore always gave the same answer, though there were cases
5998 where it should not have. (Example: a "domains = !$mx_any" option on a
5999 dnslookup router: the "domains" option is always processed without any
6000 widening, but the router might have qualify_single set.) Now Exim uses the
6001 cached value only when the same options are set.
6003 32. Added John Jetmore's "exipick" utility to the distribution.
6005 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6006 than a timeout (e.g. a certificate is required, and is not provided), an
6007 Exim server now closes the connection immediately. Previously it waited for
6008 the client to close - but if the client is SSL, it seems that they each
6009 wait for each other, leading to a delay before one of them times out.
6011 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6012 maintained 0.8.x compatibility because I don't think many are using it, and
6013 it is clearly obsolete.
6015 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6018 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6019 client certificate was expired. A simple patch fixes this, though I don't
6020 understand the full logic of why the verify callback is called multiple
6023 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6024 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6025 which causes problems with some clients (such as the Certicom SSL Plus
6026 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6027 disables the coutermeasure allowing Eudora to connect."
6029 38. Exim was not checking that a write() to a log file succeeded. This could
6030 lead to Bad Things if a log got too big, in particular if it hit a file
6031 size limit. Exim now panics and dies if it cannot write to a log file, just
6032 as it does if it cannot open a log file.
6034 39. Modified OS/Makefile-Linux so that it now contains
6036 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6038 The two -D definitions ensure that Exim is compiled with large file
6039 support, which makes it possible to handle log files that are bigger than
6042 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6043 instance) a domain was checked against a named list that involved a lookup,
6044 causing $domain_data to be set, then another domain was checked against the
6045 same list, then the first domain was re-checked, the value of $domain_data
6046 after the final check could be wrong. In particular, if the second check
6047 failed, it could be set empty. This bug probably also applied to
6050 41. The strip_trailing_dot option was not being applied to the address given
6051 with the -f command-line option.
6053 42. The code for reading a message's header from the spool was incrementing
6054 $received_count, but never initializing it. This meant that the value was
6055 incorrect (doubled) while delivering a message in the same process in which
6056 it was received. In the most common configuration of Exim, this never
6057 happens - a fresh exec is done - but it can happen when
6058 deliver_drop_privilege is set.
6060 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6061 it now includes up to 150 characters of the unexpected data in the log
6064 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6065 and building data strings. The size of both of these buffers was 10 000
6066 bytes - far larger than anybody would *ever* want, thought I. Needless to
6067 say, somebody hit the limit. I have increased the maximum line length to
6068 20 000 and the maximum data length of concatenated lines to 100 000. I have
6069 also fixed two bugs, because there was no checking on these buffers. Tsk,
6070 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6071 buffer is too small.
6073 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6074 lsearch lookups. Now it does.
6076 46. When parsing a route_list item in a manualroute router, a fixed-length
6077 buffer was used for the list of hosts. I made this 1024 bytes long,
6078 thinking that nobody would ever have a list of hosts that long. Wrong.
6079 Somebody had a whole pile of complicated expansion conditions, and the
6080 string was silently truncated, leading to an expansion error. It turns out
6081 that it is easier to change to an unlimited length (owing to other changes
6082 that have happened since this code was originally written) than to build
6083 structure for giving a limitation error. The length of the item that
6084 expands into the list of hosts is now unlimited.
6086 47. The lsearch lookup could not handle data where the length of text line was
6087 more than 4095 characters. Such lines were truncated, leading to shortened
6088 data being returned. It should now handle lines of any length.
6090 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6091 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6094 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6095 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6096 no longer happens in comment lines. A list of replacements is now placed
6097 at the head of all of the source files, except those whose only change is
6098 to replace PERL_COMMAND in the very first #! line.
6100 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6101 messages on the queue, with a bottom-up merge sort, using code contributed
6102 by Michael Haardt. This should make operations like -bp somewhat faster on
6103 large queues. It won't affect queue runners, except when queue_run_in_order
6106 51. Installed eximstats 1.31 in the distribution.
6108 52. Added support for SRV lookups to the dnslookup router.
6110 53. If an ACL referred to $message_body or $message_body_end, the value was not
6111 reset for any messages that followed in the same SMTP session.
6113 54. The store-handling optimization for building very long strings was not
6114 differentiating between the different store pools. I don't think this
6115 actually made any difference in practice, but I've tidied it.
6117 55. While running the routers to verify a sender address, $sender_address
6118 was still set to the sender address. This is wrong, because when routing to
6119 send a bounce to the sender, it would be empty. Therefore, I have changed
6120 it so that, while verifying a sender address, $sender_address is set to <>.
6121 (There is no change to what happens when verifying a recipient address.)
6123 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6124 target A or AAAA records (if not already returned) without resetting the
6125 qualify_single or search_parents options of the DNS resolver. These are
6126 inappropriate in this case because the targets of MX and SRV records must
6127 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6128 target that, when qualified, matched something in the local domain. These
6129 two options are now turned off when doing these lookups.
6131 57. It seems that at least some releases of Reiserfs (which does not have the
6132 concept of a fixed number of inodes) returns zero and not -1 for the
6133 number of available inodes. This interacted badly with check_spool_inodes,
6134 which assumed that -1 was the "no such thing" setting. What I have done is
6135 to check that the total number of inodes is greater than zero before doing
6136 the test of how many are available.
6138 58. When a "warn" ACL statement has a log_message modifier, the message is
6139 remembered, and not repeated. This is to avoid a lot of repetition when a
6140 message has many recipients that cause the same warning to be written.
6141 However, Exim was preserving the list of already written lines for an
6142 entire SMTP session, which doesn't seem right. The memory is now reset if a
6143 new message is started.
6145 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6146 debugging output unless log_rewrite was also set.
6148 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6149 of (int)(handle) when we know that handle contains (void *)(-1).
6151 61. The Exim daemon panic-logs an error return when it closes the incoming
6152 connection. However "connection reset by peer" seems to be common, and
6153 isn't really an error worthy of noting specially, so that particular error
6156 62. When Exim is trying to find all the local interfaces, it used to panic and
6157 die if the ioctl to get the interface flags failed. However, it seems that
6158 on at least one OS (Solaris 9) it is possible to have an interface that is
6159 included in the list of interfaces, but for which you get a failure error
6160 for this call. This happens when the interface is not "plumbed" into a
6161 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6162 failure of the "get flags" call assumes that the interface is down.
6164 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6165 makes life easier for people who are doing arithmetic on fields extracted
6166 from dates, where you often get leading zeros that should not be
6167 interpreted as octal.
6169 64. Added qualify_domain to the redirect router, to override the global
6172 65. If a pathologically long header line contained very many addresses (the
6173 report of this problem mentioned 10 000) and each of them was rewritten,
6174 Exim could use up a very large amount of memory. (It kept on making new
6175 copies of the header line as it rewrote, and never released the old ones.)
6176 At the expense of a bit more processing, the header rewriting function has
6177 been changed so that it no longer eats memory in this way.
6179 66. The generation of the Received: header has been moved from the time that a
6180 message starts to be received, to the time that it finishes. The timestamp
6181 in the Received: header should now be very close to that of the <= log
6182 line. There are two side-effects of this change:
6184 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6185 logged header lines no longer include the local Received: line, because
6186 it has not yet been created. The same applies to a copy of the message
6187 that is returned to a non-SMTP sender when a message is rejected.
6189 (b) When a filter file is tested using -bf, no additional Received: header
6190 is added to the test message. After some thought, I decided that this
6193 This change does not affect the value of $received_for. It is still set
6194 after address rewriting, but before local_scan() is called.
6196 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6198 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6199 gave an unhelpful panic error message, and a defer error. I have managed to
6200 change this behaviour so that it now rejects any supplied certificate,
6201 which seems right, as the list of acceptable certificates is empty.
6203 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6204 gave an unhelpful defer error. I have not managed to make this reject any
6205 supplied certificates, but the error message it gives is "no certificate
6206 supplied", which is not helpful.
6208 70. exigrep's output now also includes lines that are not associated with any
6209 message, but which match the given pattern. Implemented by a patch from
6210 Martin Sluka, which also tidied up the Perl a bit.
6212 71. Recipient callout verification, like sender verification, was using <> in
6213 the MAIL FROM command. This isn't really the right thing, since the actual
6214 sender may affect whether the remote host accepts the recipient or not. I
6215 have changed it to use the actual sender in the callout; this means that
6216 the cache record is now keyed on a recipient/sender pair, not just the
6217 recipient address. There doesn't seem to be a real danger of callout loops,
6218 since a callout by the remote host to check the sender would use <>.
6219 [SEE ABOVE: changed after hitting problems.]
6221 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6222 temporary errors. However, in the case of such a code being given after
6223 the end of a data transmission (i.e. after ".") Exim was failing to write
6224 a retry record for the message. (Yes, there was some broken host that was
6225 actually sending 8xx at this point.)
6227 73. An unknown lookup type in a host list could cause Exim to panic-die when
6228 the list was checked. (An example that provoked this was putting <; in the
6229 middle of a list instead of at the start.) If this happened during a DATA
6230 ACL check, a -D file could be left lying around. This kind of configuration
6231 error no longer causes Exim to die; instead it causes a defer error. The
6232 incident is still logged to the main and panic logs.
6234 74. Buglet left over from Exim 3 conversion. The message "too many messages
6235 in one connection" was written to the rejectlog but not the mainlog, except
6236 when address rewriting (yes!) was being logged.
6238 75. Added write_rejectlog option.
6240 76. When a system filter was run not as root (that is, when system_filter_user
6241 was set), the values of the $n variables were not being returned to the
6242 main process; thus, they were not subsequently available in the $sn
6245 77. Added +return_path_on_delivery log selector.
6247 78. A connection timeout was being treated differently from recipients deferred
6248 when testing hosts_max_try with a message that was older than the host's
6249 retry timeout. (The host should not be counted, thus allowing all hosts to
6250 be tried at least once before bouncing.) This may have been the cause of an
6251 occasionally reported bug whereby a message would remain on the queue
6252 longer than the retry timeout, but would be bounced if a delivery was
6253 forced. I say "may" because I never totally pinned down the problem;
6254 setting up timeout/retry tests is difficult. See also the next item.
6256 79. The ultimate address timeout was not being applied to errors that involved
6257 a combination of host plus message (for example, a timeout on a MAIL
6258 command). When an address resolved to a number of possible hosts, and they
6259 were not all tried for each delivery (e.g. because of hosts_max_try), a
6260 message could remain on the queue longer than the retry timeout.
6262 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6265 81. Fixed an obscure SMTP outgoing bug which required at least the following
6266 conditions: (a) there was another message waiting for the same server;
6267 (b) the server returned 5xx to all RCPT commands in the first message so
6268 that the message was not completed; (c) the server dropped the connection
6269 or gave a negative response to the RSET that Exim sends to abort the
6270 transaction. The observed case was a dropped connection after DATA that had
6271 been sent in pipelining mode. That is, the server had advertised PIPELINING
6272 but was not implementing it correctly. The effect of the bug was incorrect
6273 behaviour, such as trying another host, and this could lead to a crash.
6279 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6280 and daemon.c were passed as pointers to ints; they should have been
6281 pointers to socklen_t variables (which are typically unsigned ints).
6283 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6286 3. Fixed a really odd bug that affected only the testing scheme; patching a
6287 certain fixed string in the binary changed the value of another string that
6288 happened to be identical to the end of the original first string.
6290 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6291 name", it returns that address as the IP address. On some operating
6292 systems (e.g. Solaris), it also passes back the IP address string as the
6293 "host name". However, on others (e.g. Linux), it passes back an empty
6294 string. Exim wasn't checking for this, and was changing the host name to an
6295 empty string, assuming it had been canonicalized.
6297 5. Although rare, it is permitted to have more than one PTR record for a given
6298 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6299 all the names associated with an address, because they do in Solaris.
6300 However, it seems that they do not in Linux for data that comes from the
6301 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6302 I found this out when I moved to a new Linux workstation and tried to run
6303 the Exim test suite.
6305 To get round this problem I have changed the code so that it now does its
6306 own call to the DNS to look up PTR records when searching for a host name.
6307 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6308 addresses that are only in /etc/hosts are still found.
6310 This behaviour is, however, controlled by an option called host_lookup_
6311 order, which defaults to "bydns:byaddr". If people want to use the other
6312 order, or indeed, just use one or the other means of lookup, they can
6313 specify it in this variable.
6315 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6316 some operating systems, this comes back from gethostbyaddr() as an empty
6317 string, and this is what Exim used to test for. However, it seems that in
6318 other systems, "." is yielded. Exim now tests for this case too.
6320 7. The values of check_spool_space and check_log_space are now held internally
6321 as a number of kilobytes instead of an absolute number of bytes. If a
6322 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6323 kilobyte. This means that much larger values can be stored.
6325 8. Exim monitor: an attempt to get the action menu when not actually pointing
6326 at a message produces an empty menu entitled "No message selected". This
6327 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6328 no entries in it ("Shell widget menu has zero width and/or height"). So I
6329 have added a single, blank menu entry in this case.
6331 9. Added ${quote_local_part.
6333 10. MIME decoding is now applied to the contents of Subject: header lines when
6336 11. Now that a reference to $sender_host_address automatically causes a reverse
6337 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6338 host lookup before query-style lookups in lists that might use this
6339 variable. This has therefore been abolished, and the "net-" prefix is no
6340 longer necessary for query-style lookups.
6342 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6343 have been a typo for LFLAGS, so it has been changed.
6345 13. The install script calls Exim with "-C /dev/null" in order to find the
6346 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6347 to be output. However, since Exim outputs its version number before the
6348 error, it didn't break the script. It just looked ugly. I fixed this by
6349 always allowing "-C /dev/null" if the caller is root.
6351 14. Ignore overlarge ACL variable number when reading spool file - insurance
6352 against a later release with more variables having written the file.
6354 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6355 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6356 this, as well as the form without the "IPv6" on the front.
6358 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6359 OS/Makefile-Darwin file.
6361 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6362 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6364 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6365 variable. However, some OS use ints for inodes. I've added cast to long int
6366 to get rid of the compiler warning.
6368 19. I had forgotten to lock out "/../" in configuration file names when
6369 ALT_CONFIG_PREFIX was set.
6371 20. Routers used for verification do not need to specify transports. However,
6372 if such a router generated a host list, and callout was configured, Exim
6373 crashed, because it could not find a port number from the (non-existent)
6374 transport. It now assumes port 25 in this circumstance.
6376 21. Added the -t option to exigrep.
6378 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6379 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6380 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6381 them is set, LOOKUP_LSEARCH is forced.
6383 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6385 24. Added sender and host information to the "rejected by local_scan()" log
6386 line; previously there was no indication of these.
6388 25. Added .include_if_exists.
6390 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6391 Linux. It turns out that not all file systems support this. Apparently some
6392 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6393 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6394 not supported on the file descriptor, is now ignored when Exim is trying to
6395 sync a directory. This applies only to Linux.
6397 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6399 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6400 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6403 29. Added support for maildirsize files from supplied patch (modified a bit).
6405 30. The use of :fail: followed by an empty string could lead Exim to respond to
6406 sender verification failures with (e.g.):
6408 550 Verification failed for <xxx>
6409 550 Sender verify failed
6411 where the first response line was missing the '-' that indicates it is not
6412 the final line of the response.
6414 31. The loop for finding the name of the user that called Exim had a hardwired
6415 limit of 10; it now uses the value of finduser_retries, which is used for
6416 all other user lookups.
6418 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6421 33. Exim was neglecting to zero errno before one call of strtol() when
6422 expanding a string and expecting an integer value. On some systems this
6423 resulted in spurious "integer overflow" errors. Also, it was casting the
6424 result into an int without checking.
6426 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6427 did not work. The code looks as if it has *never* worked, though it appears
6428 to have been documented since at least release 1.62. I have made it work.
6430 35. The "timeout_DNS" error in retry rules, also documented since at least
6431 1.62, also never worked. As it isn't clear exactly what this means, and
6432 clearly it isn't a major issue, I have abolished the feature by treating it
6433 as "timeout", and writing a warning to the main and panic logs.
6435 36. The display of retry rules for -brt wasn't always showing the error code
6438 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6439 timeout_connect_A, timeout_connect_MX.
6441 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6442 to the empty sender.
6444 39. The daemon was not analysing the content of -oX till after it had closed
6445 stderr and disconnected from the controlling terminal. This meant that any
6446 syntax errors were only noted on the panic log, and the return code from
6447 the command was 0. By re-arranging the code a little, I've made the
6448 decoding happen first, so such errors now appear on stderr, and the return
6449 code is 1. However, the actual setting up of the sockets still happens in
6450 the disconnected process, so errors there are still only recorded on the
6453 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6454 connections (as happens on some IP stacks) was logged at start up time as
6455 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6456 it from "IPv6 and IPv4", which means that two separate sockets are being
6459 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6460 says whether AF_INET or AF_INET6 was passed as an argument.
6462 42. Exiwhat output was messed up when time zones were included in log
6465 43. Exiwhat now gives more information about the daemon's listening ports,
6466 and whether -tls-on-connect was used.
6468 44. The "port" option of the smtp transport is now expanded.
6470 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6471 silently ignored. Now an error message is written to the main and panic
6474 46. There's a new ACL modifier called "logwrite" which writes to a log file
6475 as soon as it is encountered.
6477 47. Added $local_user_uid and $local_user_gid at routing time.
6479 48. Exim crashed when trying to verify a sender address that was being
6482 49. Exim was recognizing only a space character after ".include". It now also
6483 recognizes a tab character.
6485 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6486 extracting the relevant information from the specification. The man page no
6487 longer contains scrambled data for the -d option, and I've added a section
6488 at the front about calling Exim under different names.
6490 51. Added "extra_headers" argument to the "mail" command in filter files.
6492 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6495 53. Installed eximstats 1.29.
6497 54. Added transport_filter_timeout as a generic transport option.
6499 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6500 Cc: header lines. This was required by RFC 822, but it not required by RFC
6503 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6504 incoming messages that did not have them. Now it does so only if the
6505 message originates locally, that is, if there is no associated remote host
6506 address. When Resent- header lines are present, this applies to the Resent-
6507 lines rather than the non-Resent- lines.
6509 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6510 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6512 58. Messages for configuration errors now include the name of the main
6513 configuration file - useful now that there may be more than one file in a
6514 list (.included file names were always shown).
6516 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6517 for those rare installations that do not start the daemon as root or run it
6518 setuid root. I've cut out the call to initgroups() if the daemon is not
6521 60. The Exim user and group can now be bound into the binary as text strings
6522 that are looked up at the start of Exim's processing.
6524 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6526 62. Added $mailstore_basename variable.
6528 63. Installed patch to sieve.c from Michael Haardt.
6530 64. When Exim failed to open the panic log after failing to open the main log,
6531 the original message it was trying to log was written to stderr and debug
6532 output, but if they were not available (the usual case in production), it
6533 was lost. Now it is written to syslog before the two lines that record the
6534 failures to open the logs.
6536 65. Users' Exim filters run in subprocesses under the user's uid. It is
6537 possible for a "deliver" command or an alias in a "personal" command to
6538 provoke an address rewrite. If logging of address rewriting is configured,
6539 this fails because the process is not running as root or exim. There may be
6540 a better way of dealing with this, but for the moment (because 4.30 needs
6541 to be released), I have disabled address rewrite logging when running a
6542 filter in a non-root, non-exim process.
6548 1. The buildconfig auxiliary program wasn't quoting the value set for
6549 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6550 not defined. This bug was masked in 4.22 by the effect that was fixed in
6553 2. Some messages that were rejected after a message id was allocated were
6554 shown as "incomplete" by exigrep. It no longer does this for messages that
6555 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6557 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6558 have allow_domain_literals set, the ID did not get logged in the <= line.
6559 Domain literals are now always recognized in Message-ID: header lines.
6561 4. The first argument for a ${extract expansion item is the key name or field
6562 number. Leading and trailing spaces in this item were not being ignored,
6563 causing some misleading effects.
6565 5. When deliver_drop_privilege was set, single queue runner processes started
6566 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6567 same command in the process it spins off) were not dropping privilege.
6569 6. When the daemon running as "exim" started a queue runner, it always
6570 re-executed Exim in the spun-off process. This is a waste of effort when
6571 deliver_drop_privilege is set. The new process now just calls the
6572 queue-runner function directly.
6578 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6581 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6582 ignored. Though the use of -oP was forcing the writing of a pid file, it
6583 was always written to the default place.
6585 3. If the message "no IP address found for host xxxx" is generated during
6586 incoming verification, it is now followed by identification of the incoming
6587 connection (so you can more easily find what provoked it).
6589 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6591 5. Added some features to "harden" Exim a bit more against certain attacks:
6593 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6594 be put in Local/Makefile. This is like the never_users runtime option,
6595 but it cannot be overridden. The default setting is "root".
6597 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6598 prefix string with which any file named in a -C command line option
6601 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6602 is retained for -C and -D only if the caller of Exim is root. Without
6603 it, the exim user may also use -C and -D and retain privilege.
6605 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6606 command line option is disabled.
6608 6. Macro names set by the -D option must start with an upper case letter, just
6609 like macro names defined in the configuration file.
6611 7. Added "dereference=" facility to LDAP.
6613 8. Two instances of the typo "uknown" in the source files are fixed.
6615 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6616 the Configure-Makefile script screwed up while processing it.
6618 10. Incorporated PCRE 4.4.
6620 11. The SMTP synchronization check was not operating right at the start of an
6621 SMTP session. For example, it could not catch a HELO sent before the client
6622 waited for the greeting. There is now a check for outstanding input at the
6623 point when the greeting is written. Because of the duplex, asynchronous
6624 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6625 way, but not yet received, when the check is performed.
6627 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6628 on TCP/IP sockets, because this apparently causes some broken clients to
6631 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6632 unchanged) from the Cygwin maintainer.
6634 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6635 was supported instead of testing for mbx. Effectively a typo.
6637 15. The spa authenticator server code was not checking that the input it
6638 received was valid base64.
6640 16. The debug output line for the "set" modifier in ACLs was not showing the
6641 name of the variable that was being set.
6643 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6645 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6646 lookup on its own. Something else was needed to trigger the lookup. For
6647 example, a match in host_lookup or the need for a host name in a host list.
6648 Now, if $sender_host_name is referenced and the host name has not yet been
6649 looked up, a lookup is performed. If the lookup fails, the variable remains
6650 empty, and $host_lookup_failed is set to "1".
6652 19. Added "eqi" as a case-independent comparison operator.
6654 20. The saslauthd authentication condition could segfault if neither service
6655 nor realm was specified.
6657 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6658 error message that was logged was misleading, and incoming SMTP
6659 connections were dropped. The message is now more accurate, and temporary
6660 errors are given to SMTP connections.
6662 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6663 (except RSET) with a 421 error, until QUIT is received. However, it was
6664 failing to send a response to QUIT.
6666 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6667 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6668 ACL is now run after the helo_try_verify_hosts code.
6670 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6671 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6672 case-independent, so other case variants are also recognized). Apparently
6673 some systems use these upper case variants.
6675 25. If more than two messages were waiting for the same host, and a transport
6676 filter was specified for the transport, Exim sent two messages over the
6677 same TCP/IP connection, and then failed with "socket operation on non-
6678 socket" when it tried to send the third.
6680 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6682 27. The extern definition of crypt16() in expand.c was not being excluded when
6683 the OS had its own crypt16() function.
6685 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6686 as a preferred synonym for return_size_limit, both as an option and as an
6689 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6691 30. Changed the default configuration ACL to relax the local part checking rule
6692 for addresses that are not in any local domains. For these addresses,
6693 slashes and pipe symbols are allowed within local parts, but the sequence
6694 /../ is explicitly forbidden.
6696 31. SPA server authentication was not clearing the challenge buffer before
6699 32. log_message in a "warn" ACL statement was writing to the reject log as
6700 well as to the main log, which contradicts the documentation and doesn't
6701 seem right (because no rejection is happening). So I have stopped it.
6703 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6704 However, I am unable to do any testing of this.
6706 34. Fixed an infelicity in the appendfile transport. When checking directories
6707 for a mailbox, to see if any needed to be created, it was accidentally
6708 using path names with one or more superfluous leading slashes; tracing
6709 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6711 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6712 contents are added to the log line that is written for every discarded
6713 recipient. (Previously a log_message setting was ignored.)
6715 36. The ${quote: operator now quotes the string if it is empty.
6717 37. The install script runs exim in order to find its version number. If for
6718 some reason other than non-existence or emptiness, which it checks, it
6719 could not run './exim', it was installing it with an empty version number,
6720 i.e. as "exim-". This error state is now caught, and the installation is
6723 38. An argument was missing from the function that creates an error message
6724 when Exim fails to connect to the socket for saslauthd authentication.
6725 This could cause Exim to crash, or give a corrupted message.
6727 39. Added isip, isip4, and isip6 to ${if conditions.
6729 40. The ACL variables $acl_xx are now saved with the message, and can be
6730 accessed later in routers, transports, and filters.
6732 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6733 strings in the file are not string-expanded.
6735 42. If a MAIL command specified a SIZE value that was too large to fit into an
6736 int variable, the check against message_size_limit failed. Such values are
6737 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6738 day this will have to be increased, but I don't think I want to be around
6739 when emails are that large.
6746 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6747 iconv() is not standard in FreeBSD.
6749 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6750 IPv6 enabled. The observed symptom was a segmentation fault on return from
6751 the function os_common_find_running_interfaces() in src/os.c.
6753 3. In the check_special_case() function in daemon.c I had used "errno" as an
6754 argument name, which causes warnings on some systems. This was basically a
6755 typo, since it was named "eno" in the comments!
6757 4. The code that waits for the clock to tick (at a resolution of some fraction
6758 of a second) so as to ensure message-id uniqueness was always waiting for
6759 at least one whole tick, when it could have waited for less. [This is
6760 almost certainly not relevant at current processor speeds, where it is
6761 unlikely to ever wait at all. But we try to future-proof.]
6763 5. The function that sleeps for a time interval that includes fractions of a
6764 second contained a race. It did not block SIGALRM between setting the
6765 timer, and suspending (a couple of lines later). If the interval was short
6766 and the sigsuspend() was delayed until after it had expired, the suspension
6767 never ended. On busy systems this could lead to processes getting stuck for
6770 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6771 process, before it forks any delivery processes. The open lookup caching
6772 mechanism meant that the open file or database connection was passed into
6773 the delivery process. The problem was that delivery processes always tidy
6774 up cached lookup data. This could cause a problem for the next delivery
6775 process started by the queue runner, because the external queue runner
6776 process does not know about the closure. So the next delivery process
6777 still has data in the lookup cache. In the case of a file lookup, there was
6778 no problem because closing a file descriptor in a subprocess doesn't affect
6779 the parent. However, if the lookup was caching a connection to a database,
6780 the connection was closed, and the second delivery process was likely to
6781 see errors such as "PGSQL: query failed: server closed the connection
6782 unexpectedly". The problem has been fixed by closing all cached lookups
6783 in a queue runner before running a delivery process.
6785 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6786 seem to have the "const" qualifier which it has on other OS. I've
6789 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6790 *of the same type* (client or server) with the same public name that an
6791 error should be diagnosed.
6793 9. When Exim looked up a host name for an IP address, but failed to find the
6794 original IP address when looking up the host name (a safety check), it
6795 output the message "<ip address> does not match any IP for NULL", which was
6796 confusing, to say the least. The bug was that the host name should have
6797 appeared instead of "NULL".
6799 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6800 user that is built into the binary, and the -C or -D options is used, root
6801 privilege is dropped before the configuration file is read. In addition,
6802 logging is switched to stderr instead of the normal log files. If the
6803 configuration then re-defines the Exim user, the unprivileged environment
6804 is probably not what is expected, so Exim logs a panic warning message (but
6807 However, if deliver_drop_privilege is set, the unprivileged state may well
6808 be exactly what is intended, so the warning has been cut out in that case,
6809 and Exim is allowed to try to write to its normal log files.
6815 1. smtp_return_error_details was not giving details for temporary sender
6816 or receiver verification errors.
6818 2. Diagnose a configuration error if two authenticators have the same public
6821 3. Exim used not to create the message log file for a message until the first
6822 delivery attempt. This could be confusing when incoming messages were held
6823 for policy or load reasons. The message log file is now created at the time
6824 the message is received, and an initial "Received" line is written to it.
6826 4. The automatically generated man page for command line options had a minor
6827 bug that caused no ill effects; however, a more serious problem was that
6828 the procedure for building the man page automatically didn't always
6829 operate. Consequently, release 4.20 contains an out-of-date version. This
6830 shouldn't happen again.
6832 5. When building Exim with embedded Perl support, the script that builds the
6833 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6834 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6836 6. The freeze_tell option was not being used for messages that were frozen on
6837 arrival, either by an ACL or by local_scan().
6839 7. Added the smtp_incomplete_transaction log selector.
6841 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6842 was accepting AUTH without a new EHLO.
6844 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6845 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6848 10. Logging of TCP/IP connections (when configured) now happens in the main
6849 daemon process instead of the child process, so that the TCP/IP connection
6850 count is more accurate (but it can never be perfect).
6852 11. The use of "drop" in a nested ACL was not being handled correctly in the
6853 outer ACL. Now, if condition failure induced by the nested "drop" causes
6854 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6855 or "require"), the connection is dropped.
6857 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6858 that yield "discard" can now be used with an "accept" or a "discard" verb,
6859 but an error is generated for any others (because I can't see a useful way
6860 to define what should happen).
6862 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6863 are now processed in the same way as lines in the Exim configuration file.
6864 In particular, continuation lines are supported.
6866 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6868 15. Added -ti meaning -t -i.
6870 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6871 domains, and warn by logging if others are found.
6873 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6874 returned from GIFCONF when Exim is trying to find the list of interfaces on
6875 a host. The code in os.c has been modified to copy each ifreq to an aligned
6876 structure in all cases.
6878 Also, in some cases, the returned ifreq's were being copied to a 'struct
6879 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6880 means the last couple of bytes of an IPv6 address could be chopped if the
6881 ifreq contained only a normal sockaddr (14 bytes storage).
6883 18. Named domain lists were not supported in the hosts_treat_as_local option.
6884 An entry such as +xxxx was not recognized, and was treated as a literal
6887 19. Ensure that header lines added by a DATA ACL are included in the reject log
6888 if the ACL subsequently rejects the message.
6890 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6891 MD5 (which is deprecated).
6893 21. When testing a filter file using -bf, Exim was writing a message when it
6894 took the sender from a "From " line in the message, but it was not doing so
6895 when it took $return_path from a Return-Path: header line. It now does.
6897 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6898 with a valid header line field name (a series of printing characters
6899 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6901 23. Changed "disc" in the source to "disk" to conform to the documentation and
6902 the book and for uniformity.
6904 24. Ignore Sendmail's -Ooption=value command line item.
6906 25. When execve() failed while trying to run a command in a pipe transport,
6907 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6908 could be confused with a return value of 69 from the command itself. This
6909 has been changed to 127, the value the shell returns if it is asked to run
6910 a non-existent command. The wording for the related log line suggests a
6911 non-existent command as the problem.
6913 26. If received_header_text expands to an empty string, do not add a Received:
6914 header line to the message. (Well, it adds a token one on the spool, but
6915 marks it "old" so that it doesn't get used or transmitted.)
6917 27. Installed eximstats 1.28 (addition of -nt option).
6919 28. There was no check for failure on the call to getsockname() in the daemon
6920 code. This can fail if there is a shortage of resources on the system, with
6921 ENOMEM, for example. A temporary error is now given on failure.
6923 29. Contrary to the C standard, it seems that in some environments, the
6924 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6925 program. Exim now does this explicitly; it affects the formatting of
6926 timestamps using strftime().
6928 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6929 complaints. I've now initialized all the variables, to avoid this.
6931 32. Header lines added by a system filter were not being "seen" during
6932 transport-time rewrites.
6934 33. The info_callback() function passed to OpenSSL is set up with type void
6935 (*)(SSL *, int, int), as described somewhere. However, when calling the
6936 function (actually a macro) that sets it up, the type void(*)() is
6937 expected. I've put in a cast to prevent warnings from picky compilers.
6939 34. If a DNS black list lookup found a CNAME record, but there were no A
6940 records associated with the domain it pointed at, Exim crashed.
6942 35. If a DNS black list lookup returned more than one A record, Exim ignored
6943 all but the first. It now scans all returned addresses if a particular IP
6944 value is being sought. In this situation, the contents of the
6945 $dnslist_value variable are a list of all the addresses, separated by a
6948 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6949 to accept a host name and all its aliases if the forward lookup for any of
6950 them yielded the IP address of the incoming connection. Now it accepts only
6951 those names whose forward lookup yields the correct IP address. Any other
6952 names are discarded. This closes a loophole whereby a rogue DNS
6953 administrator could create reverse DNS records to break through a
6954 wildcarded host restriction in an ACL.
6956 37. If a user filter or a system filter that ran in a subprocess used any of
6957 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6958 the wrong values were passed to the pipe command ($thisaddress had the
6959 value of $0, $0 had the value of $1, etc). This bug was introduced by
6960 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6962 38. Improved the line breaking for long SMTP error messages from ACLs.
6963 Previously, if there was no break point between 40 and 75 characters, Exim
6964 left the rest of the message alone. Two changes have been made: (a) I've
6965 reduced the minimum length to 35 characters; (b) if it can't find a break
6966 point between 35 and 75 characters, it looks ahead and uses the first one
6967 that it finds. This may give the occasional overlong line, but at least the
6968 remaining text gets split now.
6970 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6971 file descriptors might be low, and that setting 1000 would always raise it.
6972 It turns out that in some environments, the limit is already over 1000 and
6973 that lowering it causes trouble. So now Exim takes care not to decrease it.
6975 40. When delivering a message, the value of $return_path is set to $sender_
6976 address at the start of routing (routers may change the value). By an
6977 oversight, this default was not being set up when an address was tested by
6978 -bt or -bv, which affected the outcome if any router or filter referred to
6981 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6982 log only when "message" or "log_message" are set. However, if one of the
6983 conditions was an address verification, or a call to a nested ACL, the
6984 messages generated by the underlying test were being passed through. This
6985 no longer happens. The underlying message is available in $acl_verify_
6986 message for both "message" and "log_message" expansions, so it can be
6987 passed through if needed.
6989 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6990 new expansion $bh_ to give the encoded byte string without charset
6991 translation. Translation happens only if iconv() is available; HAVE_ICONV
6992 indicates this at build time. HEADERS_CHARSET gives the charset to
6993 translate to; headers_charset can change it in the configuration, and
6994 "headers charset" can change it in an individual filter file.
6996 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6997 that creates RFC 2047 encoded "words" labels them as that charset instead
6998 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6999 expansion operator; (ii) when Exim creates a From: line for a local
7000 message; (iii) when a header line is rewritten to include a "phrase" part.
7002 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7003 buggy, causing it to skip the first lines of messages whose message ID
7004 ended in 'D'. This would not have bitten before Exim release 4.14, because
7005 message IDs were unlikely to end in 'D' before then. The effect was to have
7006 incorrect size information for certain domains.
7008 45. #include "config.h" was missing at the start of the crypt16.c module. This
7009 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7012 46. If there was a timeout during a "random" callout check, Exim treated it as
7013 a failure of the random address, and carried on sending RSET and the real
7014 address. If the delay was just some slowness somewhere, the response to the
7015 original RCPT would be taken as a response to RSET and so on, causing
7016 mayhem of various kinds.
7018 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7019 when I implemented it. It didn't allow for the fact that some option values
7020 may legitimately be negative (e.g. size_addition), and it didn't even do
7021 the right test for positive values.
7023 48. Domain names in DNS records are case-independent. Exim always looks them up
7024 in lower case. Some resolvers return domain names in exactly the case they
7025 appear in the zone file, that is, they may contain uppercase letters. Not
7026 all resolvers do this - some return always lower case. Exim was treating a
7027 change of case by a resolver as a change of domain, similar to a widening
7028 of a domain abbreviation. This triggered its re-routing code and so it was
7029 trying to route what was effectively the same domain again. This normally
7030 caused routing to fail (because the router wouldn't handle the domain
7031 twice). Now Exim checks for this case specially, and just changes the
7032 casing of the domain that it ultimately uses when it transmits the message
7035 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7038 50. If a filter generated a file delivery with a non-absolute name (possible if
7039 no home directory exists for the router), the forbid_file option was not
7042 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7043 the existing equality matching.
7045 52. Exim was using ints instead of ino_t variables in some places where it was
7046 dealing with inode numbers.
7048 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7049 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7050 called TMPDIR, and if it finds it is different, it changes its value.
7052 54. The smtp_printf() function is now made available to local_scan() so
7053 additional output lines can be written before returning. There is also an
7054 smtp_fflush() function to enable the detection of a dropped connection.
7055 The variables smtp_input and smtp_batched_input are exported to
7058 55. Changed the default runtime configuration: the message "Unknown user"
7059 has been removed from the ACL, and instead placed on the localuser router,
7060 using the cannot_route_message feature. This means that any verification
7061 failures that generate their own messages won't get overridden. Similarly,
7062 the "Unrouteable address" message that was in the ACL for unverifiable
7063 relay addresses has also been removed.
7065 56. Added hosts_avoid_esmtp to the smtp transport.
7067 57. The exicyclog script was not checking for the esoteric option
7068 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7069 will work only if exicyclog is run under the appropriate euid.
7071 58. Following a discussion on the list, the rules by which Exim recognises line
7072 endings on incoming messages have been changed. The -dropcr and drop_cr
7073 options are now no-ops, retained only for backwards compatibility. The
7074 following line terminators are recognized: LF CRLF CR. However, special
7075 processing applies to CR:
7077 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7078 nor a local message in the state where . is a terminator.
7080 (ii) If a bare CR is encountered in a header line, an extra space is added
7081 after the line terminator so as not to end the header. The reasoning
7082 behind this is that bare CRs in header lines are most likely either
7083 to be mistakes, or people trying to play silly games.
7085 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7086 was being incorrectly given as 18 bytes larger than it should have been.
7087 This is a VOB (very old bug).
7089 60. This may never have affected anything current, but just in case it has:
7090 When the local host is found other than at the start of a list of hosts,
7091 the local host, those with the same MX, and any that follow, are discarded.
7092 When the list in question was part of a longer list of hosts, the following
7093 hosts (not currently being processed) were also being discarded. This no
7094 longer happens. I'm not sure if this situation could ever has previously
7097 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7100 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7101 header. This is something that is recommended in a new Internet Draft, and
7102 is something that is documented as being done by Sendmail. There are two
7103 possible values. For messages generated by the autoreply transport, Exim
7106 Auto-Submitted: auto-replied
7108 whereas for all other generated messages (e.g. bounces) it adds
7110 Auto-Submitted: auto-generated
7112 63. The "personal" condition in filters now includes a test for the
7113 Auto-Submitted: header. If it contains the string "auto-" the message it
7114 not considered personal.
7116 64. Added rcpt_include_affixes as a generic transport option.
7118 65. Added queue_only_override (default true).
7120 66. Added the syslog_duplication option.
7122 67. If what should have been the first header line of a message consisted of
7123 a space followed by a colon, Exim was mis-interpreting it as a header line.
7124 It isn't of course - it is syntactically invalid and should therefore be
7125 treated as the start of the message body. The misbehaviour could have
7126 caused a number of strange effects, including loss of data in subsequent
7127 header lines, and spool format errors.
7129 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7130 client host had authenticated. This control can now be exercised by an ACL
7131 for more flexibility.
7133 69. By default, callouts do not happen when testing with -bh. There is now a
7134 variant, -bhc, which does actually run the callout code, including
7135 consulting and updating the callout cache.
7137 70. Added support for saslauthd authentication, courtesy of Alexander
7140 71. If statvfs() failed on the spool or log directories while checking their
7141 size for availability, Exim confusingly gave the error "space shortage".
7142 Furthermore, in debugging mode it crashed with a floating point exception.
7143 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7144 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7145 serious problem, Exim now writes to the main and panic logs when this
7146 happens, with details of the failure. It then refuses to accept the
7147 incoming message, giving the message "spool directory problem" or "log
7148 directory problem" with a 421 code for SMTP messages.
7150 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7151 0, 1, and 2 exist, because some OS complain for execs without them (see
7152 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7153 descriptors. However, the code omitted to check that the open succeeded,
7154 causing mysterious errors if for some reason the permissions on /dev/null
7155 got screwed. Now Exim writes a message to the main and panic logs, and
7156 bombs out if it can't open /dev/null.
7158 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7159 interact so that it is all more flexible. It is supposed to remain
7160 backwards compatible. Also added extra_local_interfaces.
7162 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7163 to bomb out with an assertion failure - to the client this appears as a
7164 connection drop. This problem occurs in the part of the code that was taken
7165 from the Samba project. Fortunately, the assertion is in a very simple
7166 function, so I have fixed this by reproducing the function inline in the
7167 one place where it is called, and arranging for authentication to fail
7168 instead of killing the process with assert().
7170 75. The SPA client code was not working when the server requested OEM rather
7171 than Unicode encoding.
7173 76. Added code to make require_files with a specific uid setting more usable in
7174 the case where statting the file as root fails - usually a non-root-mounted
7175 NFS file system. When this happens and the failure is EACCES, Exim now
7176 forks a subprocess and does the per-uid checking as the relevant uid.
7178 77. Added process_log_path.
7180 78. If log_file_path was not explicitly set, a setting of check_log_space or
7181 check_log_inodes was ignored.
7183 79. If a space check for the spool or log partitions fails, the incident is now
7184 logged. Of course, in the latter case the data may get lost...
7186 80. Added the %p formatting code to string_format() so that it can be used to
7187 print addresses in debug_print(). Adjusted all the address printing in the
7188 debugging in store.c to use %p rather than %d.
7190 81. There was a concern that a line of code in smtp_in.c could overflow a
7191 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7192 initially expressed, the concern was not well-founded, because trailing
7193 spaces are removed early. However, if the trailing spaces were followed by
7194 a NULL, they did not get removed, so the overflow was possible. Two fixes
7197 (a) I re-wrote the offending code in a cleaner fashion.
7198 (b) If an incoming SMTP command contains a NULL character, it is rejected
7201 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7202 runs initgroups(), so that if the Exim user is in any additional groups,
7203 they will be used during message reception.
7209 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob