1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe funxtions being used.
21 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
22 When redis_servers is set to a list of > 1 element, and the Redis servers
23 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
24 case of MOVED into a DEFER case instead, thus moving the query onto the
25 next server in the list. For a cluster of N elements, all N servers must
26 be defined in redis_servers.
28 GF/02 Catch and remove uninitialized value warning in exiqsumm
29 Check for existence of @ARGV before looking at $ARGV[0]
31 JH/01 Replace the store_release() internal interface with store_newblock(),
32 which internalises the check required to safely use the old one, plus
33 the allocate and data copy operations duplicated in both (!) of the
36 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
37 modifier. This matches the restriction on the commandline.
39 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
40 Previously only the last row was returned.
42 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
43 we assumed that tags in the header were well-formed, and parsed the
44 element content after inspecting only the first char of the tag.
45 Assumptions at that stage could crash the receive process on malformed
48 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
49 While running the DKIM ACL we operate on the Permanent memory pool so that
50 variables created with "set" persist to the DATA ACL. Also (at any time)
51 DNS lookups that fail create cache records using the Permanent pool. But
52 expansions release any allocations made on the current pool - so a dnsdb
53 lookup expansion done in the DKIM ACL releases the memory used for the
54 DNS negative-cache, and bad things result. Solution is to switch to the
55 Main pool for expansions.
56 While we're in that code, add checks on the DNS cache during store_reset,
57 active in the testsuite.
58 Problem spotted, and debugging aided, by Wolfgang Breyha.
60 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
61 When none of the hosts presented to a transport match an already-open
62 connection, close it and proceed with the list. Previously we would
63 queue the message. Spotted by Lena with Yahoo, probably involving
66 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
67 Previously a spurious "250 OK id=" response was appended to the proper
70 JH/08 The "support for" informational output now, which built with Content
71 Scanning support, has a line for the malware scanner interfaces compiled
72 in. Interface can be individually included or not at build time.
74 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
75 by the template makefile "src/EDITME". The "STREAM" support for an older
76 ClamAV interface method is removed.
78 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
79 rows affected is given instead).
81 JH/11 The runtime Berkeley DB library version is now additionally output by
82 "exim -d -bV". Previously only the compile-time version was shown.
84 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
85 SMTP connection. Previously, when one had more receipients than the
86 first, an abortive onward connection was made. Move to full support for
87 multiple onward connections in sequence, handling cutthrough connection
88 for all multi-message initiating connections.
90 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
91 routers. Previously, a multi-recipient message would fail to match the
92 onward-connection opened for the first recipient, and cause its closure.
94 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
95 a timeout on read on a GnuTLS initiating connection, resulting in the
96 initiating connection being dropped. This mattered most when the callout
97 was marked defer_ok. Fix to keep the two timeout-detection methods
100 JH/15 Relax results from ACL control request to enable cutthrough, in
101 unsupported situations, from error to silently (except under debug)
102 ignoring. This covers use with PRDR, frozen messages, queue-only and
105 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
107 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
108 metadata, resulting in a crash in free().
110 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
111 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
112 Broken also in d185889f4, with init system revamp.
114 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
115 Previously we abruptly closed the connection after reading a malware-
116 found indication; now we go on to read the "scan ok" response line,
119 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
120 ACL. Previously, a crash would result.
122 JH/19 Speed up macro lookups during configuration file read, by skipping non-
123 macro text after a replacement (previously it was only once per line) and
124 by skipping builtin macros when searching for an uppercase lead character.
126 JH/20 DANE support moved from Experimental to mainline. The Makefile control
127 for the build is renamed.
129 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
130 was allocated for every new TLS startup, meaning one per message. Fix
131 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
133 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
134 reported the original. Fix to report (as far as possible) the ACL
135 result replacing the original.
137 JH/23 Fix memory leak during multi-message connections using STARTTLS under
138 OpenSSL. Certificate information is loaded for every new TLS startup,
139 and the resources needed to be freed.
141 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
143 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
146 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
147 DATA response info to the (existing) per-recipient response info for
148 the "C=" log element. It can have useful tracking info from the
149 destination system. Patch from Simon Arlott.
151 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
152 length value. Previously this would segfault.
154 HS/02 Support Avast multiline protoocol, this allows passing flags to
155 newer versions of the scanner.
157 JH/28 Ensure that variables possibly set during message acceptance are marked
158 dead before release of memory in the daemon loop. This stops complaints
159 about them when the debug_store option is enabled. Discovered specifically
160 for sender_rate_period, but applies to a whole set of variables.
161 Do the same for the queue-runner and queue-list loops, for variables set
162 from spool message files. Do the same for the SMTP per-message loop, for
163 certain variables indirectly set in ACL operations.
165 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
166 as a multi-recipient message from a mailinglist manager). The coding had
167 an arbitrary cutoff number of characters while checking for more input;
168 enforced by writing a NUL into the buffer. This corrupted long / fast
169 input. The problem was exposed more widely when more pipelineing of SMTP
170 responses was introduced, and one Exim system was feeding another.
171 The symptom is log complaints of SMTP syntax error (NUL chars) on the
172 receiving system, and refused recipients seen by the sending system
173 (propating to people being dropped from mailing lists).
174 Discovered and pinpointed by David Carter.
176 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
177 replaced by the ${authresults } expansion.
179 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
181 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
182 allows proper process termination in container environments.
184 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
185 Previously the "final dot" had a newline after it; ensure it is CR,LF.
187 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
188 and "err_perm", deprecated since 4.83 when the RFC-defined words
189 "temperror" and "permerror" were introduced.
191 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
192 transport-filters or DKIM-signing. The restriction was lost in the
193 consolidation of verify-callout and delivery SMTP handling.
194 Extend the restriction to also cover ARC-signing.
196 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
197 in defer=pass mode supply a 450 to the initiator. Previously the message
200 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
201 tls_require_ciphers is used as before.
203 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
204 "pass_unscanned". Only tmpfails from the scanner are written to
205 the paniclog, as they may require admin intervention (permission
206 denied, license issues). Other scanner errors (like decompression
207 bombs) do not cause a paniclog entry.
209 JH/36 Fix reinitialisation of DKIM logging variable between messages.
210 Previously it was possible to log spurious information in receive log
213 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
214 triggered odd behaviour from Outlook Express clients.
216 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
219 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
220 since the IETF WG has not yet settled on that versus the original
221 "bare" representation.
223 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
224 Previously the millisecond value corrupted the output.
225 Fix also for syslog_pid=no and log_selector +pid, for which the pid
226 corrupted the output.
232 JH/01 Rework error string handling in TLS interface so that the caller in
233 more cases is responsible for logging. This permits library-sourced
234 string to be attached to addresses during delivery, and collapses
235 pairs of long lines into single ones.
237 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
238 during configuration. Wildcards are allowed and expanded.
240 JH/02 Rework error string handling in DKIM to pass more info back to callers.
241 This permits better logging.
243 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
244 do not close it down and have the child transport start it up again on
245 the passed-on TCP connection. Instead, proxy the child (and any
246 subsequent ones) for TLS via a unix-domain socket channel. Logging is
247 affected: the continued delivery log lines do not have any DNSSEC, TLS
248 Certificate or OCSP information. TLS cipher information is still logged.
250 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
251 identical IP addresses on different listening ports. Will also affect
254 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
255 add noisy ifdef guards to special-case this sillyness.
256 Patch from Bernd Kuhls.
258 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
259 than 255 are no longer allowed.
261 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
262 Disable the session-cache too, which might reduce our load. Since we
263 currrectly use a new context for every connection, both as server and
264 client, there is no benefit for these.
265 GnuTLS appears to not support tickets server-side by default (we don't
266 call gnutls_session_ticket_enable_server()) but client side is enabled
267 by default on recent versions (3.1.3 +) unless the PFS priority string
270 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
271 <https://reproducible-builds.org/specs/source-date-epoch/>.
273 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
274 the check for any unsuccessful recipients did not notice the limit, and
275 erroneously found still-pending ones.
277 JH/08 Pipeline CHUNKING command and data together, on kernels that support
278 MSG_MORE. Only in-clear (not on TLS connections).
280 JH/09 Avoid using a temporary file during transport using dkim. Unless a
281 transport-filter is involved we can buffer the headers in memory for
282 creating the signature, and read the spool data file once for the
283 signature and again for transmission.
285 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
286 4.77 as the kernel support then wasn't solid, having issues in 64bit
287 mode. Now, it's been long enough. Add support for FreeBSD also.
289 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
290 case where the routing stage had gathered several addresses to send to
291 a host before calling the transport for the first, we previously failed
292 to close down TLS in the old transport process before passing the TCP
293 connection to the new process. The new one sent a STARTTLS command
294 which naturally failed, giving a failed delivery and bloating the retry
295 database. Investigation and fix prototype from Wolfgang Breyha.
297 JH/12 Fix check on SMTP command input synchronisation. Previously there were
298 false-negatives in the check that the sender had not preempted a response
299 or prompt from Exim (running as a server), due to that code's lack of
300 awareness of the SMTP input buffering.
302 PP/04 Add commandline_checks_require_admin option.
303 Exim drops privileges sanely, various checks such as -be aren't a
304 security problem, as long as you trust local users with access to their
305 own account. When invoked by services which pass untrusted data to
306 Exim, this might be an issue. Set this option in main configuration
307 AND make fixes to the calling application, such as using `--` to stop
310 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
311 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
312 responses to those, into a single TLS record each way (this usually means
313 a single packet). As a side issue, smtp_enforce_sync now works on TLS
316 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
317 affects you only if you're dancing at the edge of the param size limits.
318 If you are, and this message makes sense to you, then: raise the
319 configured limit or use OpenSSL 1.1. Nothing we can do for older
322 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
323 cmdline element to get the documented default one. Previously it was
326 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
327 are unsuable for non-message-specific reasons - eg. connection timeout,
328 banner-time rejection.
330 JH/16 Fix logging of delivery remote port, when specified by router, under
333 PP/06 Repair manualroute's ability to take options in any order, even if one
334 is the name of a transport.
337 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
339 JH/17 Change the list-building routines interface to use the expanding-string
340 triplet model, for better allocation and copying behaviour.
342 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
343 Previously it was constructed the first time a possibly-matching string
344 was met in the configuration file input during startup; now it is done
347 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
348 compatible one, to avoid the (poorly documented) possibility of a config
349 file in the working directory redirecting the DB files, possibly correpting
350 some existing file. CVE-2017-10140 assigned for BDB.
352 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
353 cache-hot. Previously, although the result was properly cached, the
354 initial verify call returned a defer.
356 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
357 the main verify for receipient in uncached-mode.
359 JH/22 Retire historical build files to an "unsupported" subdir. These are
360 defined as "ones for which we have no current evidence of testing".
362 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
363 if present. Previously it was ignored.
365 JH/24 Start using specified-initialisers in C structure init coding. This is
366 a C99 feature (it's 2017, so now considered safe).
368 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
369 if was a fixed-sized field and bitmask ops via macros; it is now more
372 PP/07 GitHub PR 56: Apply MariaDB build fix.
373 Patch provided by Jaroslav Škarvada.
375 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
376 during Coverity cleanups [4.87 JH/47]
377 Diagnosis and fix provided by Michael Fischer v. Mollard.
379 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
380 the right size to place the terminating semicolon on its own folded
381 line, the header hash was calculated to an incorrect value thanks to
382 the (relaxed) space the fold became.
384 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
385 and confused the parent.
387 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
388 which could crash as a result. This could lead to undeliverable messages.
390 JH/28 Logging: "next input sent too soon" now shows where input was truncated
393 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
394 matters on fast-turnover and PID-randomising systems, which were getting
395 out-of-order delivery.
397 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
398 a possibly-overlapping copy. The symptom was that "Remote host closed
399 connection in response to HELO" was logged instead of the actual 4xx
402 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
403 Previously only that bufferd was discarded, resulting in SYMTP command
406 JH/32 DKIM: when a message has multiple signatures matching an identity given
407 in dkim_verify_signers, run the dkim acl once for each. Previously only
408 one run was done. Bug 2189.
410 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
411 "panic the current process" to "deliberately defer". The panic log is
412 still written with the problem list name; the mail and reject logs now
413 get a temp-reject line for the message that was being handled, saying
414 something like "domains check lookup or other defer". The SMTP 451
415 message is still "Temporary local problem".
417 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
418 A crafted sequence of BDAT commands could result in in-use memory beeing
419 freed. CVE-2017-16943.
421 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
422 from SMTP input. Previously it was always done; now only done for DATA
423 and not BDAT commands. CVE-2017-16944.
425 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
426 to the message (such as an overlong header line). Previously this was
427 not done and we did not exit BDAT mode. Followon from the previous item
428 though a different problem.
434 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
435 than -2003 did; needs libidn2 in addition to libidn.
437 JH/02 The path option on a pipe transport is now expanded before use.
439 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
440 Patch provided by "Björn", documentation fix added too.
442 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
443 missing a wire-to-host endian conversion.
445 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
446 close after a BDAT command line could be taken as a following command,
447 giving a synch failure. Fix by only checking for synch immediately
448 before acknowledging the chunk.
450 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
451 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
452 macro. Patches provided by Josh Soref.
454 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
455 Previously we did not; the RFC seems ambiguous and VRFY is not listed
456 by IANA as a service extension. However, John Klensin suggests that we
459 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
460 the dkim code may be unix-mode line endings rather than smtp wire-format
461 CRLF, so prepend a CR to any bare LF.
463 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
464 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
466 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
467 the first were themselves being wrongly included in the feed into dkim
468 processing; with most chunk sizes in use this resulted in an incorrect
469 body hash calculated value.
471 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
472 DKIM signature block, for verification. Although advised against by
473 standards it is specifically not ruled illegal.
475 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
477 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
478 missing a body hash (the bh= tag).
480 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
481 It seems that HAProxy sends the Proxy Protocol information in clear and
482 only then does a TLS startup, so do the same.
484 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
485 TCP connections (such as for Spamd) unless the daemon successfully set
486 Fast Open mode on its listening sockets. This fixes breakage seen on
487 too-old kernels or those not configured for Fast Open, at the cost of
488 requiring both directions being enabled for TFO, and TFO never being used
489 by non-daemon-related Exim processes.
491 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
492 endings, at least on the first header line. Try to canonify any that get
493 past that check, despite the cost.
495 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
496 now limited to an arbitrary five deep, while parsing addresses with the
497 strip_excess_angle_brackets option enabled.
499 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
500 instead leave the unprompted TLS handshake in socket buffer for the
501 TLS library to consume.
503 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
505 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
507 JH/16 Drop variables when they go out of scope. Memory management drops a whole
508 region in one operation, for speed, and this leaves assigned pointers
509 dangling. Add checks run only under the testsuite which checks all
510 variables at a store-reset and panics on a dangling pointer; add code
511 explicitly nulling out all the variables discovered. Fixes one known
512 bug: a transport crash, where a dangling pointer for $sending_ip_address
513 originally assigned in a verify callout, is re-used.
515 PP/06 Drop '.' from @INC in various Perl scripts.
517 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
519 PP/08 Reduce a number of compilation warnings under clang; building with
520 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
521 should be warning-free.
523 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
525 HS/01 Fix portability problems introduced by PP/08 for platforms where
526 realloc(NULL) is not equivalent to malloc() [SunOS et al].
528 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
529 chunk. This allows us to accept broken chunked messages. We need a more
530 general solution here.
532 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
533 already-broken messages in the queue.
535 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
537 JH/19 Fix reference counting bug in routing-generated-address tracking.
543 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
544 supports it and a size is available (ie. the sending peer gave us one).
546 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
547 years of being deprecated). The replacements are the ACLs
548 acl_smtp_mime and acl_not_smtp_mime.
550 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
551 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
552 or even an in-clear connection were permitted. Now, if the host lookup
553 was dnssec and dane was requested then the host is only used if the
554 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
555 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
556 if one fails this test.
557 This means that a poorly-configured remote DNS will make it incommunicado;
558 but it protects against a DNS-interception attack on it.
560 JH/04 Bug 1810: make continued-use of an open smtp transport connection
561 non-noisy when a race steals the message being considered.
563 JH/05 If main configuration option tls_certificate is unset, generate a
564 self-signed certificate for inbound TLS connections.
566 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
567 in rewrites and routers.
569 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
570 and logged a warning sing 4.83; now they are a configuration file error.
572 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
573 (lacking @domain). Apply the same qualification processing as RCPT.
575 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
577 JH/10 Support ${sha256:} applied to a string (as well as the previous
580 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
581 a cutthrough deliver is pending, as we always want to make a connection.
582 This also avoids re-routing the message when later placing the cutthrough
583 connection after a verify cache hit.
584 Do not update it with the verify result either.
586 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
587 when routing results in more than one destination address.
589 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
590 signing (which inhibits the cutthrough capability). Previously only
591 the presence of an option was tested; now an expansion evaluating as
592 empty is permissible (obviously it should depend only on data available
593 when the cutthrough connection is made).
595 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
596 the relevant preceding SMTP command did not note the pipelining mode.
598 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
599 Previously they were not counted.
601 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
602 as one having no matching records. Previously we deferred the message
603 that needed the lookup.
605 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
606 distinguished as "(=".
608 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
609 for missing MX records. Previously it only worked for missing A records.
611 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
613 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
614 after the data-go-ahead and data-ack. Patch from Jason Betts.
616 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
617 even for a "none" policy. Patch from Tony Meyer.
619 JH/22 Fix continued use of a connection for further deliveries. If a port was
620 specified by a router, it must also match for the delivery to be
623 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
624 When one of the recipients of a message was unsuitable for the connection
625 (has no matching addresses), we lost track of needing to mark it
626 deferred. As a result mail would be lost.
628 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
630 JH/25 Decoding ACL controls is now done using a binary search; the source code
631 takes up less space and should be simpler to maintain. Merge the ACL
632 condition decode tables also, with similar effect.
634 JH/26 Fix problem with one_time used on a redirect router which returned the
635 parent address unchanged. A retry would see the parent address marked as
636 delivered, so not attempt the (identical) child. As a result mail would
639 JH/27 Fix a possible security hole, wherein a process operating with the Exim
640 UID can gain a root shell. Credit to http://www.halfdog.net/ for
641 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
644 JH/28 Enable {spool,log} filesystem space and inode checks as default.
645 Main config options check_{log,spool}_{inodes,space} are now
646 100 inodes, 10MB unless set otherwise in the configuration.
648 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
649 Previously it only applied to the main-section connection policy
652 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
654 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
655 by me. Added RFC7919 DH primes as an alternative.
657 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
658 are not in the system include path.
660 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
661 GnuTLS, when a session startup failed (eg because the client disconnected)
662 Exim did stdio operations after fclose. This was exposed by a recent
663 change which nulled out the file handle after the fclose.
665 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
666 signed directly by the cert-signing cert, rather than an intermediate
667 OCSP-signing cert. This is the model used by LetsEncrypt.
669 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
671 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
672 an incoming connection.
674 HS/02 Bug 1802: Do not half-close the connection after sending a request
677 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
678 fallback to "prime256v1".
680 JH/34 SECURITY: Use proper copy of DATA command in error message.
681 Could leak key material. Remotely exploitable. CVE-2016-9963.
687 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
688 and 3.4.4 - once the server is enabled to respond to an OCSP request
689 it does even when not requested, resulting in a stapling non-aware
690 client dropping the TLS connection.
692 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
693 support variable-length bit vectors. No functional change.
695 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
696 The I= interface field on outgoing lines is now after the H= remote
697 host field, same as incoming lines. There is a separate
698 outgoing_interface log selector which allows you to disable the
701 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
702 If not running log_selector +smtp_connection the mainlog would be held
703 open indefinitely after a "too many connections" event, including to a
704 deleted file after a log rotate. Leave the per net connection logging
705 leaving it open for efficiency as that will be quickly detected by the
706 check on the next write.
708 HS/01 Bug 1671: Fix post transport crash.
709 Processing the wait-<transport> messages could crash the delivery
710 process if the message IDs didn't exist for some reason. When
711 using 'split_spool_directory=yes' the construction of the spool
712 file name failed already, exposing the same netto behaviour.
714 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
715 mime_regex ACL conditions.
717 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
718 to DSN fail messages (bounces): remote IP, remote greeting, remote response
719 to HELO, local diagnostic string.
721 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
722 log line to debug. Even when configured with a tls authenticator many
723 client connections are expected to not authenticate in this way, so
724 an authenticate fail is not an error.
726 HS/02 Add the Exim version string to the process info. This way exiwhat
727 gives some more detail about the running daemon.
729 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
730 matter for fast-change records such as DNSBLs.
732 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
733 retry record, even if constant. There may be multiple transports with
734 different interface settings and the retry behaviour needs to be kept
737 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
739 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
741 JH/10 Bug 840: fix log_defer_output option of pipe transport
743 JH/11 Bug 830: use same host for all RCPTS of a message, even under
744 hosts_randomize. This matters a lot when combined with mua_wrapper.
746 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
747 ${quote_pgsql:<string>} operator.
749 JH/13 Bug 1708: avoid misaligned access in cached lookup.
751 JH/14 Change header file name for freeradius-client. Relevant if compiling
752 with Radius support; from the Gentoo tree and checked under Fedora.
754 JH/15 Bug 1712: Introduce $prdr_requested flag variable
756 JH/16 Bug 1714: Permit an empty string as expansion result for transport
757 option transport_filter, meaning no filtering.
759 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
761 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
762 defaults to "*" (all hosts). The variable is now available when not built
763 with TLS, default unset, mainly to enable keeping the testsuite sane.
764 If a server certificate is not supplied (via tls_certificate) an error is
765 logged, and clients will find TLS connections fail on startup. Presumably
766 they will retry in-clear.
767 Packagers of Exim are strongly encouraged to create a server certificate
768 at installation time.
770 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
771 with the $config_file variable.
773 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
774 in transport context, after the attempt, and per-recipient. The latter type
775 is per host attempted. The event data is the error message, and the errno
776 information encodes the lookup type (A vs. MX) used for the (first) host,
777 and the trailing two digits of the smtp 4xx response.
779 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
780 to write to mainlog (or rejectlog, paniclog) in the window between file
781 creation and permissions/ownership being changed. Particularly affects
782 installations where exicyclog is run as root, rather than exim user;
783 result is that the running daemon panics and dies.
785 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
787 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
788 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
789 "pri" and "weight". Note that the previous implicit priority given by the
790 list order is no longer honoured.
792 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
795 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
796 by defining SUPPORT_SOCKS.
798 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
799 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
800 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
801 variables are renamed to proxy_{local,external}_{address,port}.
803 JH/27 Move Internationalisation support from Experimental to mainline, enabled
804 for a build by defining SUPPORT_I18N
806 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
807 of the query string, and make ${quote_redis:} do that quoting.
809 JH/29 Move Events support from Experimental to mainline, enabled by default
810 and removable for a build by defining DISABLE_EVENT.
812 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
814 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
815 cached by the daemon.
817 JH/32 Move Redis support from Experimental to mainline, enabled for a build
818 by defining LOOKUP_REDIS. The libhiredis library is required.
820 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
821 keys are given for lookup.
823 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
824 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
825 only supported when built with TLS support. The PolarSSL SHA routines
826 are still used when the TLS library is too old for convenient support.
828 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
829 openssl_options), for security. OpenSSL forces this from version 1.1.0
830 server-side so match that on older versions.
832 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
833 allocation for $value could be released as the expansion processing
834 concluded, but leaving the global pointer active for it.
836 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
837 and to use the domains and local_parts ACL conditions.
839 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
840 incorrectly not doubled on cutthrough transmission, hence seen as a
841 body-termination at the receiving system - resulting in truncated mails.
842 Commonly the sender saw a TCP-level error, and retransmitted the message
843 via the normal store-and-forward channel. This could result in duplicates
844 received - but deduplicating mailstores were liable to retain only the
845 initial truncated version.
847 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
849 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
851 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
852 we're in there, support oversigning also; bug 1309.
854 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
856 HS/04 Add support for keep_environment and add_environment options.
858 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
859 either intentional arithmetic overflow during PRNG, or testing config-
862 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
863 delivery resulted in actual delivery. Cancel cutthrough before DATA
866 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
867 rejects a recipient: pass the reject to the originator.
869 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
870 Many were false-positives and ignorable, but it's worth fixing the
873 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
874 for the new environment-manipulation done at startup. Move the routines
875 from being local to tls.c to being global via the os.c file.
877 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
878 an extract embedded as result-arg for a map, the first arg for extract
879 is unavailable so we cannot tell if this is a numbered or keyed
880 extraction. Accept either.
886 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
889 JH/02 The smtp transport option "multi_domain" is now expanded.
891 JH/03 The smtp transport now requests PRDR by default, if the server offers
894 JH/04 Certificate name checking on server certificates, when exim is a client,
895 is now done by default. The transport option tls_verify_cert_hostnames
896 can be used to disable this per-host. The build option
897 EXPERIMENTAL_CERTNAMES is withdrawn.
899 JH/05 The value of the tls_verify_certificates smtp transport and main options
900 default to the word "system" to access the system default CA bundle.
901 For GnuTLS, only version 3.0.20 or later.
903 JH/06 Verification of the server certificate for a TLS connection is now tried
904 (but not required) by default. The verification status is now logged by
905 default, for both outbound TLS and client-certificate supplying inbound
908 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
911 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
912 Status Notification (bounce) messages are now MIME format per RFC 3464.
913 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
914 under the control of the dsn_advertise_hosts option, and routers may
915 have a dsn_lasthop option.
917 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
918 default, modifiable by a malware= option. The list separator for
919 the options can now be changed in the usual way. Bug 68.
921 JH/10 The smtp_receive_timeout main option is now expanded before use.
923 JH/11 The incoming_interface log option now also enables logging of the
924 local interface on delivery outgoing connections.
926 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
927 if the interface and destination host and port all match.
929 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
932 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
933 Patch from Andrew Lewis.
935 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
936 now supports optional time-restrictions, weighting, and priority
937 modifiers per server. Patch originally by <rommer@active.by>.
939 JH/16 The spamd_address main option now supports a mixed list of local
940 and remote servers. Remote servers can be IPv6 addresses, and
941 specify a port-range.
943 JH/17 Bug 68: The spamd_address main option now supports an optional
944 timeout value per server.
946 JH/18 Bug 1581: Router and transport options headers_add/remove can
947 now have the list separator specified.
949 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
952 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
955 JH/21 Support for the A6 type of dns record is withdrawn.
957 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
958 rather than the verbs used.
960 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
961 from 255 to 1024 chars.
963 JH/24 Verification callouts now attempt to use TLS by default.
965 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
966 are generic router options now. The defaults didn't change.
968 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
969 Original patch from Alexander Shikoff, worked over by JH.
971 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
972 files if ZCAT_COMMAND is not executable.
974 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
976 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
978 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
979 Normally benign, it bites when the pair was led to by a CNAME;
980 modern usage is to not canonicalize the domain to a CNAME target
981 (and we were inconsistent anyway for A-only vs AAAA+A).
983 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
985 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
986 when evaluating $sender_host_dnssec.
988 JH/31 Check the HELO verification lookup for DNSSEC, adding new
989 $sender_helo_dnssec variable.
991 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
993 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
995 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
997 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
998 documented as working, but never had. Support all but $spam_report.
1000 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1001 added for tls authenticator.
1003 HS/03 Add perl_taintmode main config option
1009 TL/01 When running the test suite, the README says that variables such as
1010 no_msglog_check are global and can be placed anywhere in a specific
1011 test's script, however it was observed that placement needed to be near
1012 the beginning for it to behave that way. Changed the runtest perl
1013 script to read through the entire script once to detect and set these
1014 variables, reset to the beginning of the script, and then run through
1015 the script parsing/test process like normal.
1017 TL/02 The BSD's have an arc4random API. One of the functions to induce
1018 adding randomness was arc4random_stir(), but it has been removed in
1019 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1020 function when detected.
1022 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1023 cause callback expansion.
1025 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1026 syntax errors in an expansion can be treated as a string instead of
1027 logging or causing an error, due to the internal use of bool_lax
1028 instead of bool when processing it.
1030 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1031 server certificates when making smtp deliveries.
1033 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1035 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1037 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1039 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1040 Merged patch from Sebastian Wiedenroth.
1042 JH/05 Fix results-pipe from transport process. Several recipients, combined
1043 with certificate use, exposed issues where response data items split
1044 over buffer boundaries were not parsed properly. This eventually
1045 resulted in duplicates being sent. This issue only became common enough
1046 to notice due to the introduction of connection certificate information,
1047 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1049 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1050 size buffer was used, resulting in syntax errors when an expansion
1053 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1054 version 3.3.6 or later.
1056 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1057 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1058 both become "event_action", the variables become $event_name, $event_data
1059 and $event_defer_errno. There is a new variable $verify_mode, usable in
1060 routers, transports and related events. The tls:cert event is now also
1061 raised for inbound connections, if the main configuration event_action
1064 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1065 early OCSP support, but no stapling (appears to be less than 1.0.0).
1067 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1068 server certificate names available under the smtp transport option
1069 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1072 JH/10 Time-related extraction expansions from certificates now use the main
1073 option "timezone" setting for output formatting, and are consistent
1074 between OpenSSL and GnuTLS compilations. Bug 1541.
1076 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1077 encoded parameter in the incoming message. Bug 1558.
1079 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1080 include certificate info, eximon was claiming there were spoolfile
1083 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1085 JH/14 Log delivery-related information more consistently, using the sequence
1086 "H=<name> [<ip>]" wherever possible.
1088 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1089 are problematic for Debian distribution, omit them from the release
1092 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1094 JH/16 Fix string representation of time values on 64bit time_t architectures.
1097 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1098 output list separator was used.
1103 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1104 checkers that were complaining about end of non-void function with no
1107 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1108 This was a regression introduced in 4.83 by another bugfix.
1110 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1112 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1113 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1119 TF/01 Correctly close the server side of TLS when forking for delivery.
1121 When a message was received over SMTP with TLS, Exim failed to clear up
1122 the incoming connection properly after forking off the child process to
1123 deliver the message. In some situations the subsequent outgoing
1124 delivery connection happened to have the same fd number as the incoming
1125 connection previously had. Exim would try to use TLS and fail, logging
1126 a "Bad file descriptor" error.
1128 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1129 utilities have not been installed.
1131 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1132 temporary space as the ACL may create new global variables.
1134 TL/01 LDAP support uses per connection or global context settings, depending
1135 upon the detected version of the libraries at build time.
1137 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1138 to extract and use the src ip:port in logging and expansions as if it
1139 were a direct connection from the outside internet. PPv2 support was
1140 updated based on HAProxy spec change in May 2014.
1142 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1144 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1145 Properly escape header and check for NULL return.
1147 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1150 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1152 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1153 characters in header names, implemented as a verify condition.
1154 Contributed by Michael Fischer v. Mollard.
1156 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1157 results permerror and temperror. Previous values are deprecated but
1158 still accepted. In a future release, err_perm and err_temp will be
1159 completely removed, which will be a backward incompatibility if the
1160 ACL tests for either of these two old results. Patch contributed by
1161 user bes-internal on the mailing list.
1163 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1165 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1166 selectors, in both main and reject logs.
1168 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1171 JH/07 Add malware type "sock" for talking to simple daemon.
1173 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1175 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1176 routers/transports under cutthrough routing.
1178 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1179 numbers. Touch up "bool" conditional to keep the same definition.
1181 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1183 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1185 JH/12 Expand items in router/transport headers_add or headers_remove lists
1186 individually rather than the list as a whole. Bug 1452.
1188 Required for reasonable handling of multiple headers_ options when
1189 they may be empty; requires that headers_remove items with embedded
1190 colons must have them doubled (or the list-separator changed).
1192 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1193 view the policy declared in the DMARC record. Currently, $dmarc_status
1194 is a combined value of both the record presence and the result of the
1197 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1199 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1200 dnslookup router and the smtp transport (applying to the forward
1203 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1204 of ldap servers used for a specific lookup. Patch provided by Heiko
1207 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1208 New variable $lookup_dnssec_authenticated for observability.
1210 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1211 Patch submitted by Lars Timman.
1213 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1215 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1216 Requires trusted mode and valid format message id, aborts otherwise.
1217 Patch contributed by Heiko Schlichting.
1219 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1220 certextract with support for various fields. Bug 1358.
1222 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1223 is requested by default, modifiable by smtp transport option
1226 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1227 operate on certificate variables to give certificate fingerprints
1228 Also new ${sha256:cert_variable}.
1230 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1232 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1233 Christian Aistleitner.
1235 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1237 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1238 file. Patch from Wolfgang Breyha.
1240 JH/25 Expand the coverage of the delivery $host and $host_address to
1241 client authenticators run in verify callout. Bug 1476.
1243 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1244 align with daemon_smtp_ports. Bug 72.
1246 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1247 support and error reporting did not work properly.
1249 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1250 and is readable. Patch from Andrew Colin Kissa.
1252 TL/14 Enhance documentation of ${run expansion and how it parses the
1253 commandline after expansion, particularly in the case when an
1254 unquoted variable expansion results in an empty value.
1256 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1258 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1259 by renaming away from it. Addresses GH issue 15, reported by
1262 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1264 TL/15 SECURITY: prevent double expansion in math comparison functions
1265 (can expand unsanitized data). Not remotely exploitable.
1272 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1274 PP/02 Make -n do something, by making it not do something.
1275 When combined with -bP, the name of an option is not output.
1277 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1280 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1281 $sender_host_name and config options to manage this, and basic check
1284 PP/05 DSCP support for outbound connections and control modifier for inbound.
1286 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1287 (Only plugin which currently uses this is kerberos4, which nobody should
1288 be using, but we should make it available and other future plugins might
1289 conceivably use it, even though it would break NAT; stuff *should* be
1290 using channel bindings instead).
1292 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1293 name; added for Sendmail compatibility; requires admin caller.
1294 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1295 just ignore it); requires trusted caller.
1296 Also parse but ignore: -Ac -Am -X<logfile>
1299 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1301 TL/02 Add +smtp_confirmation as a default logging option.
1303 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1304 Patch by Magnus Holmgren from 2007-02-20.
1306 TL/04 Bugzilla 1281 - Spec typo.
1307 Bugzilla 1283 - Spec typo.
1308 Bugzilla 1290 - Spec grammar fixes.
1310 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1312 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1314 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1315 mailing list by by Dmitry Isaikin.
1317 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1319 JH/02 Support "G" suffix to numbers in ${if comparisons.
1321 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1323 NM/01 Bugzilla 1197 - Spec typo
1324 Bugzilla 1196 - Spec examples corrections
1326 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1328 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1329 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1332 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1333 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1335 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1336 "acl {{name}{arg}...}", and optional args on acl condition
1339 JH/05 Permit multiple router/transport headers_add/remove lines.
1341 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1343 JH/07 Avoid using a waiting database for a single-message-only transport.
1344 Performance patch from Paul Fisher. Bugzilla 1262.
1346 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1349 JH/09 Add $headers_added variable, with content from use of ACL modifier
1350 add_header (but not yet added to the message). Bugzilla 199.
1352 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1353 Pulled from Bugzilla 817 by Wolfgang Breyha.
1355 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1357 (nb: this is the same fix as in Exim 4.80.1)
1359 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1362 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1365 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1367 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1368 advertise SMTP AUTH mechanism to us, instead of a generic
1369 protocol violation error. Also, make Exim more robust to bad
1370 data from the Dovecot auth socket.
1372 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1374 When a queue runner is handling a message, Exim first routes the
1375 recipient addresses, during which it prunes them based on the retry
1376 hints database. After that it attempts to deliver the message to
1377 any remaining recipients. It then updates the hints database using
1380 So if a recipient address works intermittently, it can get repeatedly
1381 deferred at routing time. The retry hints record remains fresh so the
1382 address never reaches the final cutoff time.
1384 This is a fairly common occurrence when a user is bumping up against
1385 their storage quota. Exim had some logic in its local delivery code
1386 to deal with this. However it did not apply to per-recipient defers
1387 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1389 This change adds a proper retry rule check during routing so that the
1390 final cutoff time is checked against the message's age. We only do
1391 this check if there is an address retry record and there is not a
1392 domain retry record; this implies that previous attempts to handle
1393 the address had the retry_use_local_parts option turned on. We use
1394 this as an approximation for the destination being like a local
1395 delivery, as in LMTP.
1397 I suspect this new check makes the old local delivery cutoff check
1398 redundant, but I have not verified this so I left the code in place.
1400 TF/02 Correct gecos expansion when From: is a prefix of the username.
1402 Test 0254 submits a message to Exim with the header
1406 When I ran the test suite under the user fanf2, Exim expanded
1407 the header to contain my full name, whereas it should have added
1408 a Resent-Sender: header. It erroneously treats any prefix of the
1409 username as equal to the username.
1411 This change corrects that bug.
1413 GF/01 DCC debug and logging tidyup
1414 Error conditions log to paniclog rather than rejectlog.
1415 Debug lines prefixed by "DCC: " to remove any ambiguity.
1417 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1419 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1420 Bug spotted by Jeremy Harris; was flawed since initial commit.
1421 Would have resulted in OCSP responses post-SNI triggering an Exim
1422 NULL dereference and crash.
1424 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1426 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1427 Bug detection, analysis and fix by Samuel Thibault.
1428 Bugzilla 1331, Debian bug #698092.
1430 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1432 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1433 Server implementation by Todd Lyons, client by JH.
1434 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1435 config variable "prdr_enable" controls whether the server
1436 advertises the facility. If the client requests PRDR a new
1437 acl_data_smtp_prdr ACL is called once for each recipient, after
1438 the body content is received and before the acl_smtp_data ACL.
1439 The client is controlled by both of: a hosts_try_prdr option
1440 on the smtp transport, and the server advertisement.
1441 Default client logging of deliveries and rejections involving
1442 PRDR are flagged with the string "PRDR".
1444 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1445 fclose(). Diagnosis by Todd Lyons.
1447 PP/17 Update configure.default to handle IPv6 localhost better.
1448 Patch by Alain Williams (plus minor tweaks).
1451 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1452 This is now consistent with GnuTLS, and is now documented: the
1453 previous undocumented portable approach to treating the option as
1454 unset was to force an expansion failure. That still works, and
1455 an empty string is now equivalent.
1457 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1458 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1459 not performing validation itself.
1461 PP/20 Added force_command boolean option to pipe transport.
1462 Patch from Nick Koston, of cPanel Inc.
1464 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1467 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1469 PP/21 Fix eximon continuous updating with timestamped log-files.
1470 Broken in a format-string cleanup in 4.80, missed when I repaired the
1471 other false fix of the same issue.
1472 Report and fix from Heiko Schlichting.
1475 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1476 Report from Prashanth Katuri.
1478 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1479 It's SecureTransport, so affects any MacOS clients which use the
1480 system-integrated TLS libraries, including email clients.
1482 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1483 using a MIME ACL for non-SMTP local injection.
1484 Report and assistance in diagnosis by Warren Baker.
1486 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1488 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1490 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1491 last id that failed so it may be referenced in subsequent ACL's.
1493 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1496 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1497 ldap library initialization, allowing self-signed CA's to be
1498 used. Also properly sets require_cert option later in code by
1499 using NULL (global ldap config) instead of ldap handle (per
1500 session). Bug diagnosis and testing by alxgomz.
1502 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1503 the src/util/ subdirectory.
1505 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1506 renamed to Transport Post Delivery Action by Jeremy Harris, as
1509 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1510 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1511 redis_servers = needs to be configured which will be used by the redis
1512 lookup. Patch from Warren Baker, of The Packet Hub.
1514 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1516 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1517 hostname or reverse DNS when processing a host list. Used suggestions
1518 from multiple comments on this bug.
1520 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1522 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1523 Missed a few lines, added it to make the runtest require no keyboard
1526 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1527 contains upper case chars. Make router use caseful_local_part.
1529 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1530 support when GnuTLS has been built with p11-kit.
1536 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1538 This, or similar/improved, will also be change PP/11 of 4.82.
1544 PP/01 Handle short writes when writing local log-files.
1545 In practice, only affects FreeBSD (8 onwards).
1546 Bugzilla 1053, with thanks to Dmitry Isaikin.
1548 NM/01 Bugzilla 949 - Documentation tweak
1550 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1553 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1555 PP/02 Implemented gsasl authenticator.
1557 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1559 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1560 `pkg-config foo` for cflags/libs.
1562 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1563 with rest of GSASL and with heimdal_gssapi.
1565 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1566 `pkg-config foo` for cflags/libs for the TLS implementation.
1568 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1569 properties get this fed in as external SSF. A number of robustness
1570 and debugging improvements to the cyrus_sasl authenticator.
1572 PP/08 cyrus_sasl server now expands the server_realm option.
1574 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1575 Patch by Jeremy Harris.
1577 PP/10 Added dbmjz lookup type.
1579 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1581 PP/12 MAIL args handles TAB as well as SP, for better interop with
1582 non-compliant senders.
1583 Analysis and variant patch by Todd Lyons.
1585 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1586 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1587 Patch from Dirk Mueller <dmueller@suse.com>
1589 PP/13 tls_peerdn now print-escaped for spool files.
1590 Observed some $tls_peerdn in wild which contained \n, which resulted
1591 in spool file corruption.
1593 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1594 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1595 or write after TLS renegotiation, which otherwise led to messages
1598 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1599 as a tracking header (ie: a signed header comes before the signature).
1600 Patch from Wolfgang Breyha.
1602 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1603 comma-sep list; embedded commas doubled.
1605 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1607 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1609 Report and patch from Dmitry Banschikov.
1611 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1612 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1613 failures. We appear to now support TLS1.1+ with Exim.
1615 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1616 lets Exim select keys and certificates based upon TLS SNI from client.
1617 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1618 before an outbound SMTP session. New log_selector, +tls_sni.
1620 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1621 NULL dereference. Report and patch from Alun Jones.
1623 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1624 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1625 Not seeing resolver debug output on NetBSD, but suspect this is a
1626 resolver implementation change.
1628 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1629 Left warnings. Added "eximon gdb" invocation mode.
1631 PP/21 Defaulting "accept_8bitmime" to true, not false.
1633 PP/22 Added -bw for inetd wait mode support.
1635 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1636 locate the relevant includes and libraries. Made this the default.
1638 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1639 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1641 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1642 This may cause build issues on older platforms.
1644 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1645 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1646 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1647 Added SNI support via GnuTLS too.
1648 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1650 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1652 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1653 Applied second patch from Janne, implementing suggestion to default
1654 multiple-strings-in-record handling to match SPF spec.
1656 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1658 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1659 read-only, out of scope).
1660 Patch from Wolfgang Breyha, report from Stuart Northfield.
1662 PP/29 Fix three issues highlighted by clang analyser static analysis.
1663 Only crash-plausible issue would require the Cambridge-specific
1664 iplookup router and a misconfiguration.
1665 Report from Marcin Mirosław.
1667 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1669 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1670 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1671 As part of this, removing so much warning spew let me fix some minor
1672 real issues in debug logging.
1674 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1675 assignment on my part. Fixed.
1677 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1678 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1679 Janne Snabb (who went above and beyond: thank you).
1681 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1682 string otherwise requires a connection and a bunch more work and it's
1683 relatively easy to get wrong. Should also expose TLS library linkage
1686 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1687 64-bit ${eval} (JH/03).
1689 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1690 GNU libc to support some of the 64-bit stuff, should not lead to
1691 conflicts. Defined before os.h is pulled in, so if a given platform
1692 needs to override this, it can.
1694 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1695 protection layer was required, which is not implemented.
1696 Bugzilla 1254, patch from Wolfgang Breyha.
1698 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1699 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1700 tls_dhparam take prime identifiers. Also unbreak combination of
1701 OpenSSL+DH_params+TLSSNI.
1703 PP/39 Disable SSLv2 by default in OpenSSL support.
1709 PP/01 Solaris build fix for Oracle's LDAP libraries.
1710 Bugzilla 1109, patch from Stephen Usher.
1712 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1714 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1717 TF/02 Fix a couple more cases where we did not log the error message
1718 when unlink() failed. See also change 4.74-TF/03.
1720 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1721 lock up or crash if it happened to be inside a call to libc when it
1722 got a SIGUSR1 from exiwhat.
1724 The SIGUSR1 handler appends the current process status to the process
1725 log which is later printed by exiwhat. It used to use the general
1726 purpose logging code to do this, but several functions it calls are
1727 not safe for signals.
1729 The new output code in the SIGUSR1 handler is specific to the process
1730 log, and simple enough that it's easy to inspect for signal safety.
1731 Removing some special cases also simplifies the general logging code.
1732 Removing the spurious timestamps from the process log simplifies
1735 TF/04 Improved ratelimit ACL condition.
1737 The /noupdate option has been deprecated in favour of /readonly which
1738 has clearer semantics. The /leaky, /strict, and /readonly update modes
1739 are mutually exclusive. The update mode is no longer included in the
1740 database key; it just determines when the database is updated. (This
1741 means that when you upgrade Exim will forget old rate measurements.)
1743 Exim now checks that the per_* options are used with an update mode that
1744 makes sense for the current ACL. For example, when Exim is processing a
1745 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1746 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1747 must specify per_mail/readonly. If you omit the update mode it defaults to
1748 /leaky where that makes sense (as before) or /readonly where required.
1750 The /noupdate option is now undocumented but still supported for
1751 backwards compatibility. It is equivalent to /readonly except that in
1752 ACLs where /readonly is required you may specify /leaky/noupdate or
1753 /strict/noupdate which are treated the same as /readonly.
1755 A useful new feature is the /count= option. This is a generalization
1756 of the per_byte option, so that you can measure the throughput of other
1757 aggregate values. For example, the per_byte option is now equivalent
1758 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1760 The per_rcpt option has been generalized using the /count= mechanism
1761 (though it's more complicated than the per_byte equivalence). When it is
1762 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1763 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1764 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1765 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1766 Note that using per_rcpt with a non-readonly update mode in more than
1767 one ACL will cause the recipients to be double-counted. (The per_mail
1768 and per_byte options don't have this problem.)
1770 The handling of very low rates has changed slightly. If the computed rate
1771 is less than the event's count (usually one) then this event is the first
1772 after a long gap. In this case the rate is set to the same as this event's
1773 count, so that the first message of a spam run is counted properly.
1775 The major new feature is a mechanism for counting the rate of unique
1776 events. The new per_addr option counts the number of different
1777 recipients that someone has sent messages to in the last time period. It
1778 behaves like per_rcpt if all the recipient addresses are different, but
1779 duplicate recipient addresses do not increase the measured rate. Like
1780 the /count= option this is a general mechanism, so the per_addr option
1781 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1782 example, measure the rate that a client uses different sender addresses
1783 with the options per_mail/unique=$sender_address. There are further
1784 details in the main documentation.
1786 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1788 TF/06 Removed a few PCRE remnants.
1790 TF/07 Automatically extract Exim's version number from tags in the git
1791 repository when doing development or release builds.
1793 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1794 Bugzilla 879. Patch from Paul Fisher.
1796 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1797 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1800 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1802 PP/05 Variable $av_failed, true if the AV scanner deferred.
1803 Bugzilla 1078. Patch from John Horne.
1805 PP/06 Stop make process more reliably on build failure.
1806 Bugzilla 1087. Patch from Heiko Schlittermann.
1808 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1809 Bugzilla 1089. Patch from Heiko Schlittermann.
1811 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1812 Bugzilla 1131. Patch from Holger Weiß.
1814 PP/09 Handle IPv6 addresses with SPF.
1815 Bugzilla 860. Patch from Wolfgang Breyha.
1817 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1819 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1822 PP/11 match_* no longer expand right-hand-side by default.
1823 New compile-time build option, EXPAND_LISTMATCH_RHS.
1824 New expansion conditions, "inlist", "inlisti".
1826 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1828 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1830 PP/14 fix log_write() format string regression from TF/03.
1831 Bugzilla 1152. Patch from Dmitry Isaikin.
1837 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1839 PP/02 Harmonised TLS library version reporting; only show if debugging.
1840 Layout now matches that introduced for other libraries in 4.74 PP/03.
1842 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1844 PP/04 New "dns_use_edns0" global option.
1846 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1849 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1850 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1852 TK/01 Updated PolarSSL code to 0.14.2.
1853 Bugzilla 1097. Patch from Andreas Metzler.
1855 PP/07 Catch divide-by-zero in ${eval:...}.
1856 Fixes bugzilla 1102.
1858 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1861 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1862 format-string attack -- SECURITY: remote arbitrary code execution.
1864 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1865 time unintentionally subject to list matching rules, letting the header
1866 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1867 arbitrary string expansion). This allowed for information disclosure.
1869 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1870 INT_MIN/-1 -- value coerced to INT_MAX.
1876 NM/01 Workaround for PCRE version dependency in version reporting
1879 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1880 This fixes portability to compilers other than gcc, notably
1881 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1883 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1884 makefiles for portability to HP-UX and POSIX correctness.
1886 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1887 Also via indented variable definition in the Makefile.
1888 (Debugging by Oliver Heesakkers).
1890 PP/02 Restore caching of spamd results with expanded spamd_address.
1891 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1893 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1894 Improves build reliability. Fix from: Frank Elsner
1896 NM/02 Fix wide character breakage in the rfc2047 coding
1897 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1899 NM/03 Allow underscore in dnslist lookups
1900 Fixes bug 1026. Patch from Graeme Fowler
1902 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1903 Code patches from Adam Ciarcinski of NetBSD.
1905 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1908 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1909 is logged, to avoid truncation. Patch from John Horne.
1911 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1912 Patch from Jakob Hirsch.
1914 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1915 SQL string expansion failure details.
1916 Patch from Andrey Oktyabrski.
1918 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1919 Patch from Simon Arlott.
1921 PP/09 New lookups functionality failed to compile on old gcc which rejects
1922 extern declarations in function scope.
1923 Patch from Oliver Fleischmann
1925 PP/10 Use sig_atomic_t for flags set from signal handlers.
1926 Check getgroups() return and improve debugging.
1927 Fixed developed for diagnosis in bug 927 (which turned out to be
1930 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1931 Patch from Mark Zealey.
1933 PP/12 Bugzilla 1056: Improved spamd server selection.
1934 Patch from Mark Zealey.
1936 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1937 Based on patch from Heiko Schlittermann.
1939 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1940 Patch from Uwe Doering, sign-off by Michael Haardt.
1942 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1943 variable declaration deep within a block. Bug and patch from
1946 PP/15 lookups-Makefile IRIX compatibility coercion.
1948 PP/16 Make DISABLE_DKIM build knob functional.
1950 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1951 Patch by Simon Arlott
1953 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1954 variable argument macros. Our copy now differs from upstream.
1960 TF/01 Failure to get a lock on a hints database can have serious
1961 consequences so log it to the panic log.
1963 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1964 controlled using the smtp_confirmation log selector.
1966 TF/03 Include the error message when we fail to unlink a spool file.
1968 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1969 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1970 for maintaining out-of-tree patches for some time.
1972 PP/01 Bugzilla 139: Documentation and portability issues.
1973 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1974 Handle per-OS dynamic-module compilation flags.
1976 PP/02 Let /dev/null have normal permissions.
1977 The 4.73 fixes were a little too stringent and complained about the
1978 permissions on /dev/null. Exempt it from some checks.
1979 Reported by Andreas M. Kirchwitz.
1981 PP/03 Report version information for many libraries, including
1982 Exim version information for dynamically loaded libraries. Created
1983 version.h, now support a version extension string for distributors
1984 who patch heavily. Dynamic module ABI change.
1986 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1987 privilege escalation vulnerability whereby the Exim run-time user
1988 can cause root to append content of the attacker's choosing to
1991 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1994 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1995 If dropping privileges for untrusted macros, we disabled normal logging
1996 on the basis that it would fail; for the Exim run-time user, this is not
1997 the case, and it resulted in successful deliveries going unlogged.
1998 Fixed. Reported by Andreas Metzler.
2004 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2005 only prepend for the Resent-* case. Fixes regression introduced in
2006 Exim 4.70 by NM/22 for Bugzilla 607.
2008 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2009 increasing numbers of administrators be bitten by this.
2011 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2013 PP/03 Bugzilla 994: added openssl_options main configuration option.
2015 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2017 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2019 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2021 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2022 without a peer certificate, leading to a segfault because of an
2023 assumption that peers always have certificates. Be a little more
2024 paranoid. Problem reported by Martin Tscholak.
2026 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2027 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2028 NB: ClamAV planning to remove STREAM in "middle of 2010".
2029 CL also introduces -bmalware, various -d+acl logging additions and
2030 more caution in buffer sizes.
2032 PP/09 Implemented reverse_ip expansion operator.
2034 PP/10 Bugzilla 937: provide a "debug" ACL control.
2036 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2038 PP/12 Bugzilla 973: Implement --version.
2040 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2042 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2044 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2046 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2047 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2048 ignore trailing whitespace.
2050 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2052 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2055 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2056 Notification from Dr Andrew Aitchison.
2058 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2059 ExtendedDetectionInfo response format.
2060 Notification from John Horne.
2062 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2065 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2066 XSL and documented dependency on system catalogs, with examples of how
2069 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2072 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2073 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2074 configuration file which is writeable by the Exim user or group.
2076 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2077 of configuration files to cover files specified with the -C option if
2078 they are going to be used with root privileges, not just the default
2081 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2082 option (effectively making it always true).
2084 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2085 files to be used while preserving root privileges.
2087 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2088 that rogue child processes cannot use them.
2090 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2091 run-time user, instead of root.
2093 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2094 Exim run-time user without dropping privileges.
2096 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2097 result string, instead of calling string_vformat() twice with the same
2100 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2101 for other users. Others should always drop root privileges if they use
2102 -C on the command line, even for a whitelisted configure file.
2104 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2106 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2112 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2113 $data_path, and $header_path variables; fixed documentation bugs and
2116 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2117 exipick to access non-standard spools, including the "frozen" queue
2120 NM/01 Bugzilla 965: Support mysql stored procedures.
2121 Patch from Alain Williams
2123 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2125 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2126 Patch from Andreas Metzler
2128 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2129 Patch from Kirill Miazine
2131 NM/05 Bugzilla 671: Added umask to procmail example.
2133 JJ/03 installed exipick 20100323.0, fixing doc bug
2135 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2136 directory. Notification and patch from Dan Rosenberg.
2138 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2140 TK/02 Improve log output when DKIM signing operation fails.
2142 MH/01 Treat the transport option dkim_domain as a colon separated
2143 list, not as a single string, and sign the message with each element,
2144 omitting multiple occurences of the same signer.
2146 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2147 Bugzilla 985, 986. Patch by Simon Arlott
2149 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2150 Patch by Simon Arlott
2152 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2153 MBX locking. Notification from Dan Rosenberg.
2159 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2161 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2163 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2165 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2167 NM/04 Bugzilla 915: Flag broken perl installation during build.
2173 TK/01 Added patch by Johannes Berg that expands the main option
2174 "spamd_address" if it starts with a dollar sign.
2176 TK/02 Write list of recipients to X-Envelope-Sender header when building
2177 the mbox-format spool file for content scanning (suggested by Jakob
2180 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2181 (http://www.dcc-servers.net/) support via dccifd. Activated by
2182 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2184 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2185 by Mark Daniel Reidel <mr@df.eu>.
2187 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2188 When building exim an external PCRE library is now needed -
2189 PCRE is a system library on the majority of modern systems.
2190 See entry on PCRE_LIBS in EDITME file.
2192 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2193 conversation. Added nologin parameter to request.
2194 Patch contributed by Kirill Miazine.
2196 TF/01 Do not log submission mode rewrites if they do not change the address.
2198 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2200 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2201 log files in place. Contributed by Roberto Lima.
2203 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2205 TF/03 Bugzilla 615: When checking the local_parts router precondition
2206 after a local_part_suffix or local_part_prefix option, Exim now
2207 does not use the address's named list lookup cache, since this
2208 contains cached lookups for the whole local part.
2210 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2211 Robert Millan. Documentation is in experimental-spec.txt.
2213 TF/04 Bugzilla 668: Fix parallel build (make -j).
2215 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2217 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2218 Patch provided by Jan Srzednicki.
2220 TF/05 Leading white space used to be stripped from $spam_report which
2221 wrecked the formatting. Now it is preserved.
2223 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2224 that they are available at delivery time.
2226 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2228 TF/08 TLS error reporting now respects the incoming_interface and
2229 incoming_port log selectors.
2231 TF/09 Produce a more useful error message if an SMTP transport's hosts
2232 setting expands to an empty string.
2234 NM/06 Bugzilla 744: EXPN did not work under TLS.
2235 Patch provided by Phil Pennock.
2237 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2238 Patch provided by Richard Godbee.
2240 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2241 acl_smtp_notquit, added index entry.
2243 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2244 Patch provided by Eugene Bujak.
2246 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2247 accept(). Patch provided by Maxim Dounin.
2249 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2250 Patch provided by Phil Pennock.
2252 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2254 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2255 Patch provided by Brad "anomie" Jorsch.
2257 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2258 Patch provided by Dean Brooks.
2260 TK/05 Add native DKIM support (does not depend on external libraries).
2262 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2263 Patch provided by Graeme Fowler.
2265 NM/16 Bugzilla 851: Documentation example syntax fix.
2267 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2269 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2272 NM/19 Bugzilla 745: TLS version reporting.
2273 Patch provided by Phil Pennock.
2275 NM/20 Bugzilla 167: bool: condition support.
2276 Patch provided by Phil Pennock.
2278 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2279 clients. Patch provided by Phil Pennock.
2281 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2282 Patch provided by Brad "anomie" Jorsch.
2284 NM/23 Bugzilla 687: Fix misparses in eximstats.
2285 Patch provided by Heiko Schlittermann.
2287 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2288 Patch provided by Heiko Schlittermann.
2290 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2291 plus update to original patch.
2293 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2295 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2296 Patch provided by David Brownlee.
2298 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2300 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2302 NM/30 Bugzilla 888: TLS documentation bugfixes.
2304 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2306 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2307 Unlike the original bugzilla I have changed all shell scripts in src tree.
2309 NM/33 Bugzilla 898: Transport filter timeout fix.
2310 Patch by Todd Rinaldo.
2312 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2313 Patch by Serge Demonchaux.
2315 NM/35 Bugzilla 39: Base64 decode bug fixes.
2316 Patch by Jakob Hirsch.
2318 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2320 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2322 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2324 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2330 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2331 ALT-N's libdkim that I have put here:
2332 http://duncanthrax.net/exim-experimental/
2334 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2335 were called 'true' and it seems that C99 defines that as a reserved
2336 keyword to be used with 'bool' variable types. That means you could
2337 not include C99-style headers which use bools without triggering
2338 build errors in sieve.c.
2340 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2341 as mailq or other aliases. Changed the --help handling significantly
2342 to do whats expected. exim_usage() emits usage/help information.
2344 SC/01 Added the -bylocaldomain option to eximstats.
2346 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2348 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2350 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2356 PH/01 Another patch from the Sieve maintainer.
2358 PH/02 When an IPv6 address is converted to a string for single-key lookup
2359 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2360 dots are used instead of colons so that keys in lsearch files need not
2361 contain colons. This was done some time before quoting was made available
2362 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2363 (notated using the quote facility) so as to distinguish them from IPv4
2364 keys. This meant that lookups for IP addresses in host lists did not work
2365 for iplsearch lookups.
2367 This has been fixed by arranging for IPv6 addresses to be expressed with
2368 colons if the lookup type is iplsearch. This is not incompatible, because
2369 previously such lookups could never work.
2371 The situation is now rather anomalous, since one *can* have colons in
2372 ordinary lsearch keys. However, making the change in all cases is
2373 incompatible and would probably break a number of configurations.
2375 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2378 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2379 conversion specification without a maximum field width, thereby enabling
2380 a rogue spamd server to cause a buffer overflow. While nobody in their
2381 right mind would setup Exim to query an untrusted spamd server, an
2382 attacker that gains access to a server running spamd could potentially
2383 exploit this vulnerability to run arbitrary code as the Exim user.
2385 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2386 $primary_hostname instead of what libspf2 thinks the hosts name is.
2388 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2389 a directory entry by the name of the lookup key. Previously, if a
2390 symlink pointed to a non-existing file or a file in a directory that
2391 Exim lacked permissions to read, a lookup for a key matching that
2392 symlink would fail. Now it is enough that a matching directory entry
2393 exists, symlink or not. (Bugzilla 503.)
2395 PH/03 The body_linecount and body_zerocount variables are now exported in the
2398 PH/04 Added the $dnslist_matched variable.
2400 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2401 This means they are set thereafter only if the connection becomes
2404 PH/06 Added the client_condition to authenticators so that some can be skipped
2405 by clients under certain conditions.
2407 PH/07 The error message for a badly-placed control=no_multiline_responses left
2408 "_responses" off the end of the name.
2410 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2412 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2413 (without spaces) instead of just copying the configuration text.
2415 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2417 PH/11 Added $max_received_linelength.
2419 PH/12 Added +ignore_defer and +include_defer to host lists.
2421 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2422 way in which PCRE > 7.0 is built.
2424 PH/14 Implemented queue_only_load_latch.
2426 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2427 MAIL command. The effect was to mangle the value on 64-bit systems.
2429 PH/16 Another patch from the Sieve maintainer.
2431 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2433 PH/18 If a system quota error occurred while trying to create the file for
2434 a maildir delivery, the message "Mailbox is full" was not appended to the
2435 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2436 only to a quota excession during the actual writing of the file.
2438 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2439 characters?) which causes problems in log lines. The DN values are now
2440 passed through string_printing() before being added to log lines.
2442 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2443 and InterBase are left for another time.)
2445 PH/21 Added message_body_newlines option.
2447 PH/22 Guard against possible overflow in moan_check_errorcopy().
2449 PH/23 POSIX allows open() to be a macro; guard against that.
2451 PH/24 If the recipient of an error message contained an @ in the local part
2452 (suitably quoted, of course), incorrect values were put in $domain and
2453 $local_part during the evaluation of errors_copy.
2459 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2460 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2463 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2464 issue a MAIL command.
2466 PH/02 In an ACL statement such as
2468 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2470 if a client was not listed at all, or was listed with a value other than
2471 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2472 the condition was not true (as it should be), so access was not denied.
2473 The bug was that the ! inversion was incorrectly passed on to the second
2474 item. This has been fixed.
2476 PH/03 Added additional dnslists conditions == and =& which are different from
2477 = and & when the dns lookup returns more than one IP address.
2479 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2480 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2482 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2483 FSYNC, which compiles an option called disable_fsync that allows for
2484 bypassing fsync(). The documentation is heavily laced with warnings.
2486 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2488 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2489 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2490 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2491 including adding "make clean"; (3) Added -fPIC when compiling the test
2492 dynamically loaded module, to get rid of a warning.
2494 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2495 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2496 = 0s. The bug is otherwise harmless.
2498 PH/07 There was a bug in the dovecot authenticator such that the value of
2499 $auth1 could be overwritten, and so not correctly preserved, after a
2500 successful authentication. This usually meant that the value preserved by
2501 the server_setid option was incorrect.
2503 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2505 PH/09 Installed PCRE release 7.0.
2507 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2508 run for batched SMTP input. It is now run at the start of every message
2509 in the batch. While fixing this I discovered that the process information
2510 (output by running exiwhat) was not always getting set for -bs and -bS
2511 input. This is fixed, and it now also says "batched" for BSMTP.
2513 PH/11 Added control=no_pipelining.
2515 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2516 patch, slightly modified), and move the expansion of helo_data till after
2517 the connection is made in the smtp transport (so it can use these
2520 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2522 PH/14 Added log_selector = +pid.
2524 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2526 PH/16 Add ${if forany and ${if forall.
2528 PH/17 Added dsn_from option to vary the From: line in DSNs.
2530 PH/18 Flush SMTP output before performing a callout, unless control =
2531 no_callout_flush is set.
2533 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2534 was true (the default) a successful delivery failed to delete the retry
2535 item, thus causing premature timeout of the address. The bug is now
2538 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2540 PH/21 Long custom messages for fakedefer and fakereject are now split up
2541 into multiline responses in the same way that messages for "deny" and
2542 other ACL rejections are.
2544 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2545 with slight modification.
2547 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2548 draft, changing the syntax and factoring some duplicate code".
2550 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2551 for deliveries of the second and subsequent messages over the same SMTP
2554 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2555 ${reduce, with only minor "tidies".
2557 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2559 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2560 expansion side effects.
2562 PH/27 When a message times out after an over-quota error from an Exim-imposed
2563 quota, the bounce message says "mailbox is full". This message was not
2564 being given when it was a system quota that was exceeded. It now should
2567 MH/03 Made $recipients available in local_scan(). local_scan() already has
2568 better access to the recipient list through recipients_list[], but
2569 $recipients can be useful in postmaster-provided expansion strings.
2571 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2572 in the case of a MAIL command with additional options following the
2573 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2574 were accidentally chopped off.
2576 PH/29 SMTP synchronization checks are implemented when a command is read -
2577 there is a check that no more input is waiting when there shouldn't be
2578 any. However, for some commands, a delay in an ACL can mean that it is
2579 some time before the response is written. In this time, more input might
2580 arrive, invalidly. So now there are extra checks after an ACL has run for
2581 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2582 pipelining has not been advertised.
2584 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2586 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2587 returned in the SMTP response when smtp_return_error_details was false.
2588 This has been fixed.
2590 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2591 instead of the C I/O that was originally supplied, because problems were
2592 reported on Solaris.
2594 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2595 Exim which did not show up earlier: it was assuming that a call to
2596 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2597 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2598 was a macro that became an assignment, so it seemed to work. This has
2599 changed to a proper function call with a void return, hence the compile
2600 error. Exim's code has been fixed.
2602 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2605 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2607 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2609 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2610 because it turns out to be too big or there is a timeout) there is no
2611 "Completed" line in the log. When some messages of this type were
2612 selected by exigrep, they were listed as "not completed". Others were
2613 picked up by some special patterns. I have improved the selection
2614 criteria to be more general.
2616 PH/38 The host_find_failed option in the manualroute router can now be set
2617 to "ignore", to completely ignore a host whose IP address cannot be
2618 found. If all hosts are ignored, the behaviour is controlled by the new
2619 host_all_ignored option.
2621 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2622 homing or because of multiple MX records with /mx) generated more than
2623 one IP address, and the following item turned out to be the local host,
2624 all the secondary addresses of the first item were incorrectly removed
2625 from the list, along with the local host and any following hosts (which
2626 is what is supposed to happen).
2628 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2629 whoever called Exim into the -H file. In the case of the daemon it was
2630 behaving confusingly. When first started, it used values for whoever
2631 started the daemon, but after a SIGHUP it used the Exim user (because it
2632 calls itself on a restart). I have changed the code so that it now always
2635 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2636 message are rejected with the same error (e.g. no authentication or bad
2637 sender address), and a DATA command is nevertheless sent (as can happen
2638 with PIPELINING or a stupid MUA), the error message that was given to the
2639 RCPT commands is included in the rejection of the DATA command. This is
2640 intended to be helpful for MUAs that show only the final error to their
2643 PH/42 Another patch from the Sieve maintainer.
2645 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2646 Eximstats - Fixed some broken HTML links and added missing column headers
2648 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2649 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2651 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2653 PH/43 Yet another patch from the Sieve maintainer.
2655 PH/44 I found a way to check for a TCP/IP connection going away before sending
2656 the response to the final '.' that terminates a message, but only in the
2657 case where the client has not sent further data following the '.'
2658 (unfortunately, this is allowed). However, in many cases there won't be
2659 any further data because there won't be any more messages to send. A call
2660 to select() can be used: if it shows that the input is "ready", there is
2661 either input waiting, or the socket has been closed. An attempt to read
2662 the next input character can distinguish the two cases. Previously, Exim
2663 would have sent an OK response which the client would never have see.
2664 This could lead to message repetition. This fix should cure that, at
2665 least in a lot of common cases.
2667 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2668 advertised in response to EHLO.
2674 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2675 fixed by 4.65/MH/01 (is this a record?) are fixed:
2677 (i) An empty string was always treated as zero by the numeric comparison
2678 operators. This behaviour has been restored.
2680 (ii) It is documented that the numeric comparison operators always treat
2681 their arguments as decimal numbers. This was broken in that numbers
2682 starting with 0 were being interpreted as octal.
2684 While fixing these problems I realized that there was another issue that
2685 hadn't been noticed. Values of message_size_limit (both the global option
2686 and the transport option) were treated as octal if they started with 0.
2687 The documentation was vague. These values are now always treated as
2688 decimal, and I will make that clear in the documentation.
2694 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2695 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2698 MH/01 Don't check that the operands of numeric comparison operators are
2699 integers when their expansion is in "skipping" mode (fixes bug
2700 introduced by 4.64-PH/07).
2702 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2703 child addresses, Exim now panics and dies. Previously, because the count
2704 is held in a short int, deliveries were likely to be lost. As such a
2705 large number of recipients for a single message is ridiculous
2706 (performance will be very, very poor), I have chosen to impose a limit
2707 rather than extend the field.
2713 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2714 leftover -K file (the existence of which was triggered by #402).
2715 While we were at it, introduced process PID as part of the -K
2716 filename. This should rule out race conditions when creating
2719 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2720 processing considerably. Previous code took too long for large mails,
2721 triggering a timeout which in turn triggers #401.
2723 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2724 in the DK code in transports.c. sendfile() is not really portable,
2725 hence the _LINUX specificness.
2727 TF/01 In the add_headers option to the mail command in an Exim filter,
2728 there was a bug that Exim would claim a syntax error in any
2729 header after the first one which had an odd number of characters
2732 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2733 callout verification, Exim cached a "reject" for the entire domain. This
2734 is correct for most verifications, but it is not correct for a recipient
2735 verification with use_sender or use_postmaster set, because in that case
2736 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2737 case of MAIL FROM:<> rejection from other early rejections (e.g.
2738 rejection of HELO). When verifying a recipient using a non-null MAIL
2739 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2740 Whatever the result of the callout, the value of the domain cache is
2741 left unchanged (for any other kind of callout, getting as far as trying
2742 RCPT means that the domain itself is ok).
2744 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2747 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2748 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2749 session, the daemon ignores ECONNECT errors and logs others; it now
2750 ignores EPIPE as well.
2752 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2753 (quoted-printable decoding).
2755 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2756 later the small subsequent patch to fix an introduced bug.
2758 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2760 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2762 PH/08 An error is now given if message_size_limit is specified negative.
2764 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2765 to be given (somewhat) arbitrary names.
2767 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2770 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2771 miscellaneous code fixes
2773 PH/10 Added the log_reject_target ACL modifier to specify where to log
2776 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2777 hostname. This is wrong, because it relates to the incoming message (and
2778 probably the interface on which it is arriving) and not to the outgoing
2779 callout (which could be using a different interface). This has been
2780 changed to use the value of the helo_data option from the smtp transport
2781 instead - this is what is used when a message is actually being sent. If
2782 there is no remote transport (possible with a router that sets up host
2783 addresses), $smtp_active_hostname is used.
2785 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2786 tweaks were necessary in order to get it to work (see also 21 below):
2787 (a) The code assumed that strncpy() returns a negative number on buffer
2788 overflow, which isn't the case. Replaced with Exim's string_format()
2790 (b) There were several signed/unsigned issues. I just did the minimum
2791 hacking in of casts. There is scope for a larger refactoring.
2792 (c) The code used strcasecmp() which is not a standard C function.
2793 Replaced with Exim's strcmpic() function.
2794 (d) The code set only $1; it now sets $auth1 as well.
2795 (e) A simple test gave the error "authentication client didn't specify
2796 service in request". It would seem that Dovecot has changed its
2797 interface. Fortunately there's a specification; I followed it and
2798 changed what the client sends and it appears to be working now.
2800 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2803 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2804 address A is aliased to B and C, where B exists and C does not. Without
2805 -v the output is "A verified" because verification stops after a
2806 successful redirection if more than one address is generated. However,
2807 with -v the child addresses are also verified. Exim was outputting "A
2808 failed to verify" and then showing the successful verification for C,
2809 with its parentage. It now outputs "B failed to verify", showing B's
2810 parentage before showing the successful verification of C.
2812 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2813 look up a TXT record in a specific list after matching in a combined
2816 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2817 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2818 they consult the DNS. I had assumed they would set it the way they
2819 wanted; and indeed my experiments on Linux seem to show that in some
2820 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2821 To be on the safe side, however, I have now made the interface to
2822 host_find_byname() similar to host_find_bydns(), with an argument
2823 containing the DNS resolver options. The host_find_byname() function now
2824 sets these options at its start, just as host_find_bydns() does. The smtp
2825 transport options dns_qualify_single and dns_search_parents are passed to
2826 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2827 of host_find_byname() use the default settings of RES_DEFNAMES
2828 (qualify_single) but not RES_DNSRCH (search_parents).
2830 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2831 spool_read_header() do less string testing, by means of a preliminary
2832 switch on the second character of optional "-foo" lines. (This is
2833 overdue, caused by the large number of possibilities that now exist.
2834 Originally there were few.) While I was there, I also converted the
2835 str(n)cmp tests so they don't re-test the leading "-" and the first
2836 character, in the hope this might squeeze out yet more improvement.
2838 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2839 flag allowing group syntax was set by the header_syntax check but not
2840 turned off, possible causing trouble later; (2) The flag was not being
2841 set at all for the header_verify test, causing "group"-style headers to
2842 be rejected. I have now set it in this case, and also caused header_
2843 verify to ignore an empty address taken from a group. While doing this, I
2844 came across some other cases where the code for allowing group syntax
2845 while scanning a header line wasn't quite right (mostly, not resetting
2846 the flag correctly in the right place). These bugs could have caused
2847 trouble for malformed header lines. I hope it is now all correct.
2849 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2850 with the "reply" argument non-NULL. The code, however (which originally
2851 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2852 but it didn't always do it. This confused somebody who was copying the
2853 code for some other use. I have removed all the tests.
2855 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2856 feature that was used to support insecure browsers during the U.S. crypto
2857 embargo. It requires special client support, and Exim is probably the
2858 only MTA that supported it -- and would never use it because real RSA is
2859 always available. This code has been removed, because it had the bad
2860 effect of slowing Exim down by computing (never used) parameters for the
2861 RSA_EXPORT functionality.
2863 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2864 authenticator to fail if there's a tab character in the incoming data
2865 (there should never be unless someone is messing about, as it's supposed
2866 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2867 if the connection is using TLS or if the remote IP is the same as the
2868 local IP, and the "valid-client-cert option" if a client certificate has
2871 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2872 authenticators. This can be used for authorization after authentication
2873 succeeds. (In the case of plaintext, it servers for both authentication
2876 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2877 if any retry times were supplied.
2879 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2880 connection, where there is no client IP address to check. In this
2881 situation, the verify now always succeeds.
2883 PH/25 Applied John Jetmore's -Mset patch.
2885 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2887 PH/27 In a string expansion for a processed (not raw) header when multiple
2888 headers of the same name were present, leading whitespace was being
2889 removed from all of them, but trailing whitespace was being removed only
2890 from the last one. Now trailing whitespace is removed from each header
2891 before concatenation. Completely empty headers in a concatenation (as
2892 before) are ignored.
2894 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2895 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2897 PH/29 [Removed. This was a change that I later backed out, and forgot to
2898 correct the ChangeLog entry (that I had efficiently created) before
2899 committing the later change.]
2901 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2902 address errors (4xx response to RCPT) over the same connection as other
2903 messages routed to the same hosts. Such deliveries are always "forced",
2904 so retry times are not inspected. This resulted in far too many retries
2905 for the affected addresses. The effect occurred only when there were more
2906 hosts than the hosts_max_try setting in the smtp transport when it had
2907 the 4xx errors. Those hosts that it had tried were not added to the list
2908 of hosts for which the message was waiting, so if all were tried, there
2909 was no problem. Two fixes have been applied:
2911 (i) If there are any address or message errors in an SMTP delivery, none
2912 of the hosts (tried or untried) are now added to the list of hosts
2913 for which the message is waiting, so the message should not be a
2914 candidate for sending over the same connection that was used for a
2915 successful delivery of some other message. This seems entirely
2916 reasonable: after all the message is NOT "waiting for some host".
2917 This is so "obvious" that I'm not sure why it wasn't done
2918 previously. Hope I haven't missed anything, but it can't do any
2919 harm, as the worst effect is to miss an optimization.
2921 (ii) If, despite (i), such a delivery is accidentally attempted, the
2922 routing retry time is respected, so at least it doesn't keep
2923 hammering the server.
2925 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2926 in ${readsocket because some servers need this prod.
2928 PH/32 Added some extra debug output when updating a wait-xxx database.
2930 PH/33 The hint "could be header name not terminated by colon", which has been
2931 given for certain expansion errors for a long time, was not being given
2932 for the ${if def:h_colon_omitted{... case.
2934 PH/34 The spec says: "With one important exception, whenever a domain list is
2935 being scanned, $domain contains the subject domain." There was at least
2936 one case where this was not true.
2938 PH/35 The error "getsockname() failed: connection reset by peer" was being
2939 written to the panic log as well as the main log, but it isn't really
2940 panic-worthy as it just means the connection died rather early on. I have
2941 removed the panic log writing for the ECONNRESET error when getsockname()
2944 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2945 runs only) independently of the message's sender address. This meant
2946 that, if the 4xx error was in fact related to the sender, a different
2947 message to the same recipient with a different sender could confuse
2948 things. In particular, this can happen when sending to a greylisting
2949 server, but other circumstances could also provoke similar problems.
2950 I have changed the default so that the retry time for these errors is now
2951 based a combination of the sender and recipient addresses. This change
2952 can be overridden by setting address_retry_include_sender=false in the
2955 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2956 remote server are returned as part of bounce messages. This was not
2957 happening for LMTP over a pipe (the lmtp transport), but now it is the
2958 same for both kinds of LMTP.
2960 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2961 in header lines that were in fact CNAMEs. This is no longer the case.
2963 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2964 and queue runs started by the daemon processed all messages. This has
2965 been fixed so that -R and -S can now usefully be given with -q<time>.
2967 PH/40 Import PCRE release 6.7 (fixes some bugs).
2969 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2971 PH/42 Give an error if -q is specified more than once.
2973 PH/43 Renamed the variables $interface_address and $interface_port as
2974 $received_ip_address and $received_port, to make it clear that these
2975 values apply to message reception, and not to the outgoing interface when
2976 a message is delivered. (The old names remain recognized, of course.)
2978 PH/44 There was no timeout on the connect() call when using a Unix domain
2979 socket in the ${readsocket expansion. There now is.
2981 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2982 be meaningful with "accept".
2984 SC/01 Eximstats V1.43
2985 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2987 SC/02 Eximstats V1.44
2988 Use a glob alias rather than an array ref in the generated
2989 parser. This improves both readability and performance.
2991 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2992 Collect SpamAssassin and rejection statistics.
2993 Don't display local sender or destination tables unless
2994 there is data to show.
2995 Added average volumes into the top table text output.
2997 SC/04 Eximstats V1.46
2998 Collect data on the number of addresses (recipients)
2999 as well as the number of messages.
3001 SC/05 Eximstats V1.47
3002 Added 'Message too big' to the list of mail rejection
3003 reasons (thanks to Marco Gaiarin).
3005 SC/06 Eximstats V1.48
3006 Mainlog lines which have GMT offsets and are too short to
3007 have a flag are now skipped.
3009 SC/07 Eximstats V1.49 (Alain Williams)
3010 Added the -emptyok flag.
3012 SC/08 Eximstats V1.50
3013 Fixes for obtaining the IP address from reject messages.
3015 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3016 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3017 whitespace changes from 4.64-PH/27
3019 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3022 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3023 are found, allow negative numbers in numeric criteria)
3025 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3027 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3028 to match changes made in 4.64-PH/43
3030 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3032 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3033 "domain" is actually an IP address, and force a failure. This locks out
3034 those revolvers/nameservers that support "A-for-A" lookups, in
3035 contravention of the specifications.
3037 PH/48 When a host name was looked up from an IP address, and the subsequent
3038 forward lookup of the name timed out, the host name was left in
3039 $sender_host_name, contrary to the specification.
3041 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3042 restricted to single-key lookups, Exim was not diagnosing an error if
3043 * or *@ was used with a query-style lookup.
3045 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3047 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3048 long ago, but noone interested enough thought of it. Let's just say that
3049 the "1.1" means that there are some new functions that weren't there at
3050 some point in the past.
3052 PH/51 Error processing for expansion failure of helo_data from an smtp
3053 transport during callout processing was broken.
3055 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3056 tested/used via the -bh/-bhc/-bs options.
3058 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3059 bug, fixed in subsequent PCRE releases).
3061 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3062 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3064 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3070 SC/01 Use a glob alias rather than an array ref in eximstats generated
3071 parser. This improves both readability and performance.
3073 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3074 Don't display local sender or destination tables in eximstats unless
3075 there is data to show.
3076 Added average volumes into the eximstats top table text output.
3078 SC/03 Collect data on the number of addresses (recipients) as well
3079 as the number of messages in eximstats.
3081 TF/01 Correct an error in the documentation for the redirect router. Exim
3082 does (usually) call initgroups() when daemonizing.
3084 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3085 with consistent privilege compared to when running as a daemon.
3087 TF/03 Note in the spec that $authenticated_id is not set for local
3088 submissions from trusted users.
3090 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3091 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3093 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3094 by adding some example configuration directives to the default
3095 configuration file. A little bit of work is required to uncomment the
3096 directives and define how usernames and passwords are checked, but
3097 there is now a framework to start from.
3099 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3100 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3101 without this. I don't know how relevant this is to other LDAP libraries.
3103 PH/02 Add the verb name to the "unknown ACL verb" error.
3105 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3107 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3109 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3110 directory not expanded when it should be if an expanded home directory
3111 was set for the address (which is overridden by the transport).
3113 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3116 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3117 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3118 because it is too late at that time, and has no effect.
3120 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3121 security issue with \' (bugzilla #107). I could not use the
3122 PQescapeStringConn() function, because it needs a PGconn value as one of
3125 PH/08 When testing addresses using -bt, indicate those final addresses that
3126 are duplicates that would not cause an additional delivery. At least one
3127 person was confused, thinking that -bt output corresponded to deliveries.
3128 (Suppressing duplicates isn't a good idea as you lose the information
3129 about possibly different redirections that led to the duplicates.)
3131 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3132 systems where poll() doesn't work, in particular OS X.
3134 PH/10 Added more information to debugging output for retry time not reached.
3136 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3137 operations in malware.c.
3139 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3142 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3143 syslog_duplication set false, log lines that would normally be written
3144 both the the main log and to the reject log were not written to syslog at
3147 PH/14 In the default configuration, change the use of "message" in ACL warn
3148 statements to "add_header".
3150 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3151 not followed by a command (e.g. "seen endif").
3153 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3154 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3157 PH/17 Added extra conditions to the default value of delay_warning_condition
3161 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3162 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3163 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3166 The Auto-Submitted: and various List- headers are standardised, whereas I
3167 don't think Precedence: ever was.
3169 PH/18 Refactored debugging code in route_finduser() to show more information,
3170 in particular, the error code if getpwnam() issues one.
3172 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3173 This is apparently needed in addition to the PH/07 change above to avoid
3174 any possible encoding problems.
3176 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3177 but not after initializing Perl.
3179 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3180 output them only if debugging. By default they are written stderr,
3181 apparently, which is not desirable.
3183 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3186 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3189 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3191 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3192 authenticated or an ident call has been made. Suppress the default
3193 values for $authenticated_id and $authenticated_sender (but permit -oMai
3194 and -oMas) when testing with -bh.
3196 PH/24 Re-jigged the order of the tests in the default configuration so that the
3197 tests for valid domains and recipients precede the DNS black list and CSA
3198 tests, on the grounds that those ones are more expensive.
3200 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3201 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3202 This bug exists in every version of Exim that I still have, right back to
3205 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3206 However, an attempt to turn on case-sensitivity in a regex key by
3207 including (?-i) didn't work because the subject string was already
3208 lowercased, and the effects were non-intuitive. It turns out that a
3209 one-line patch can be used to allow (?-i) to work as expected.
3215 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3216 other effects) broke the use of negated acl sub-conditions.
3218 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3221 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3222 "Deny" causes Exim to reject the incoming connection with a 554 error.
3223 Unfortunately, if there is a major crisis, such as a disk failure,
3224 tcp-wrappers gives "deny", whereas what one would like would be some
3225 kind of temporary error. A kludge has been added to help with this.
3226 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3227 554 error is used if errno is still zero or contains ENOENT (which occurs
3228 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3231 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3233 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3234 errors. Otherwise a message that provokes a temporary error (when other
3235 messages do not) can cause a whole host to time out.
3237 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3238 addresses were routed directly to files or pipes from a redirect router.
3239 File deliveries just didn't batch; pipe deliveries might have suffered
3242 PH/06 A failure to get a lock for a hints database would erroneously always say
3243 "Failed to get write lock", even when it was really a read lock.
3245 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3246 of 0600. This has been changed to use the value of the lockfile_mode
3247 option (which defaults to 0600).
3249 PH/08 Applied small patch from the Sieve maintainer.
3251 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3252 folder from quota calculations, a direct delivery into this folder messed
3253 up the contents of the maildirsize file. This was because the regex was
3254 used only to exclude .Trash (or whatever) when the size of the mailbox
3255 was calculated. There was no check that a delivery was happening into an
3256 excluded directory. This bug has been fixed by ignoring all quota
3257 processing for deliveries into excluded directories.
3259 PH/10 Added the maildirfolder_create_regex option to appendfile.
3265 PH/01 The code for finding all the local interface addresses on a FreeBSD
3266 system running IPv6 was broken. This may well have applied to all BSD
3267 systems, as well as to others that have similar system calls. The broken
3268 code found IPv4 interfaces correctly, but gave incorrect values for the
3269 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3270 that it would not match correctly against @[] and not recognize the IPv6
3273 PH/02 The ipliteral router was not recognizing addresses of the form user@
3274 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3276 PH/03 Added disable_ipv6.
3278 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3279 decoded form, because it is most often used to construct To: headers
3280 lines in autoreplies, and the decoded form may well be syntactically
3281 invalid. However, $reply_address has leading white space removed, and all
3282 newlines turned into spaces so that the autoreply transport does not
3285 PH/05 If group was specified without a user on a router, and no group or user
3286 was specified on a transport, the group from the router was ignored.
3288 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3289 for visible compile-time settings that can be used to change these
3290 numbers, for those that want even more. Backwards compatibility with old
3291 spool files has been maintained. However, going back to a previous Exim
3292 release will lost any variables that are in spool files.
3294 PH/07 Two small changes when running in the test harness: increase delay when
3295 passing a TCP/IP connection to a new process, in case the original
3296 process has to generate a bounce, and remove special handling of
3297 127.0.0.2 (sic), which is no longer necessary.
3299 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3300 be the same on different OS.
3302 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3305 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3306 whether --show-vars was specified or not
3308 JJ/02 exipick: Added support for new ACL variable spool format introduced
3311 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3312 syntactically invalid From: or Reply-to: line, and a filter used this to
3313 generate an autoreply, and therefore failed to obtain an address for the
3314 autoreply, Exim could try to deliver to a non-existent relative file
3315 name, causing unrelated and misleading errors. What now happens is that
3316 it logs this as a hard delivery error, but does not attempt to create a
3319 PH/11 The exinext utility has a -C option for testing purposes, but although
3320 the given file was scanned by exinext itself; it wasn't being passed on
3321 when Exim was called.
3323 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3324 an end-of-file indication when reading a command response.
3326 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3327 compiled. In many other places in Exim, IPv6 addresses are always
3328 recognized, so I have changed this. It also means that IPv4 domain
3329 literals of the form [IPV4:n.n.n.n] are now always recognized.
3331 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3332 used if the router is not running as root, for example, when verifying at
3333 ACL time, or when using -bh. The debugging output from this situation was
3334 non-existent - all you got was a failure to exec. I have made two
3337 (a) Failures to set uid/gid, the current directory, or a process leader
3338 in a subprocess such as that created by queryprogram now generate
3339 suitable debugging output when -d is set.
3341 (b) The queryprogram router detects when it is not running as root,
3342 outputs suitable debugging information if -d is set, and then runs
3343 the subprocess without attempting to change uid/gid.
3345 PH/15 Minor change to Makefile for building test_host (undocumented testing
3348 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3349 additional section of a DNS packet that returns MX or SRV records.
3350 Instead, it always explicitly searches for A/AAAA records. This avoids
3351 major problems that occur when a DNS server includes only records of one
3352 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3353 fixed another bug: if SRV records were looked up and the corresponding
3354 address records were *not* found in the additional section, the port
3355 values from the SRV records were lost.
3357 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3358 using the correct key (the original address) when searching the retry
3359 rules in order to find which one to use for generating the retry hint.
3361 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3362 adding the default one. Similarly, if it contains a Reply-To: header, the
3363 errors_reply_to option, if set, is not used.
3365 PH/19 When calculating a retry time, Exim used to measure the "time since
3366 failure" by looking at the "first failed" field in the retry record. Now
3367 it does not use this if it is later than than the arrival time of the
3368 message. Instead it uses the arrival time. This makes for better
3369 behaviour in cases where some deliveries succeed, thus re-setting the
3370 "first failed" field. An example is a quota failure for a huge message
3371 when small messages continue to be delivered. Without this change, the
3372 "time since failure" will always be short, possible causing more frequent
3373 delivery attempts for the huge message than are intended.
3374 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3376 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3377 $1, $2, $3) because the numerical variables can be reset during some
3378 expansion items (e.g. "match"), thereby losing the authentication data.
3380 PH/21 Make -bV show the size of off_t variables so that the test suite can
3381 decide whether to run tests for quotas > 2G.
3383 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3384 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3385 filecount value is greater than 2G or if a quota value is greater than 2G
3386 on a system where the size of off_t is not greater than 4, a panic error
3389 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3390 never match. The debug and -bh output now contains an explicit error
3391 message indicating a malformed IPv4 address or mask.
3393 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3394 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3395 PH/23 above applies.
3397 PH/25 Do not write to syslog when running in the test harness. The only
3398 occasion when this arises is a failure to open the main or panic logs
3399 (for which there is an explicit test).
3401 PH/26 Added the /no_tell option to "control=freeze".
3403 PH/27 If a host name lookup failed very early in a connection, for example, if
3404 the IP address matched host_lookup and the reverse lookup yielded a name
3405 that did not have a forward lookup, an error message of the form "no IP
3406 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3407 could be logged. Now it outputs the IP address instead of "NULL".
3409 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3410 allows the sender and the authenticated sender to be set when
3411 submitting a message from within Exim. Since child_open_exim() is
3412 documented for local_scan(), the new function should be too.
3414 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3415 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3416 results in an empty string is now treated as unset.
3418 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3420 PH/31 Added sender_verify_fail logging option.
3422 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3423 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3424 tidied the source and removed it altogether.
3426 PH/33 When a queue run was abandoned because the load average was too high, a
3427 log line was always written; now it is written only if the queue_run log
3428 selector is set. In addition, the log line for abandonment now contains
3429 information about the queue run such as the pid. This is always present
3430 in "start" and "stop" lines but was omitted from the "abandon" line.
3432 PH/34 Omit spaces between a header name and the colon in the error message that
3433 is given when verify = headers_syntax fails (if there are lots of them,
3434 the message gets confusing).
3436 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3437 names, as there are now some PTR records that contain slashes. This check
3438 is only to protect against broken name servers that fall over on strange
3439 characters, so the fact that it applies to all lookups doesn't matter.
3441 PH/36 Now that the new test suite is complete, we can remove some of the
3442 special code in Exim that was needed for the old test suite. For example,
3443 sorting DNS records because real resolvers return them in an arbitrary
3444 order. The new test suite's fake resolver always returns records in the
3447 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3448 bounces) except when queue_only is set, to avoid logging races between
3449 the different processes.
3451 PH/38 Panic-die if .include specifies a non-absolute path.
3453 PH/39 A tweak to the "H" retry rule from its user.
3455 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3456 a label. They prevented compilation on older perls.
3458 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3459 a warning to be raised on newish perls.
3461 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3462 on queue. Changes to match documented behaviour of showing count of
3463 messages matching specified criteria.
3465 PH/40 Changed the default ident timeout from 30s to 5s.
3467 PH/41 Added support for the use of login_cap features, on those BSD systems
3468 that have them, for controlling the resources used by pipe deliveries.
3470 PH/42 The content-scanning code uses fopen() to create files in which to put
3471 message data. Previously it was not paying any attention to the mode of
3472 the files. Exim runs with umask(0) because the rest of the code creates
3473 files with open(), and sets the required mode explicitly. Thus, these
3474 files were ending up world-writeable. This was not a big issue, because,
3475 being within the spool directory, they were not world-accessible. I have
3476 created a function called modefopen, which takes an additional mode
3477 argument. It sets umask(777), creates the file, chmods it to the required
3478 mode, then resets the umask. All the relevant calls to fopen() in the
3479 content scanning code have been changed to use this function.
3481 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3482 to 24 hours. This avoids potential overflow problems when processing G
3483 and H retry rules. I suspect nobody ever tinkers with this value.
3485 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3487 PH/45 When the plaintext authenticator is running as a client, the server's
3488 challenges are checked to ensure they are valid base64 strings. By
3489 default, the authentication attempt is cancelled if an invalid string is
3490 received. Setting client_ignore_invalid_base64 true ignores these errors.
3491 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3492 they are received. Thus, the responses can be made to depend on the
3493 challenges. If an invalid string is ignored, an empty string is placed in
3496 PH/46 Messages that are created by the autoreply transport now contains a
3497 References: header, in accordance with RFCs 2822 and 3834.
3499 PH/47 Added authenticated_sender_force to the smtp transport.
3501 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3503 PH/49 Installed latest patch from the Sieve maintainer.
3505 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3506 was also set, the appendfile transport was unnecessarily scanning a
3507 directory of message files (e.g. for maildir delivery) to find the count
3508 of files (along with the size), even though it did not need this
3509 information. It now does the scan only if it needs to find either the
3510 size of the count of files.
3512 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3514 PH/52 Two bugs concerned with error handling when the smtp transport is
3517 (i) Exim was not creating retry information for temporary errors given
3518 for individual recipients after the DATA command when the smtp transport
3519 was used in LMTP mode. This meant that they could be retried too
3520 frequently, and not timed out correctly.
3522 (ii) Exim was setting the flag that allows error details to be returned
3523 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3524 recipients that were returned after the DATA command.
3526 PH/53 This is related to PH/52, but is more general: for any failing address,
3527 when detailed error information was permitted to be returned to the
3528 sender, but the error was temporary, then after the final timeout, only
3529 "retry timeout exceeded" was returned. Now it returns the full error as
3530 well as "retry timeout exceeded".
3532 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3533 do this, and (what is worse) MTAs that accept it.
3535 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3536 will now be deprecated.
3538 PH/56 New os.c-cygwin from the Cygwin maintainer.
3540 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3541 formats (previously only available in exim formats via -bpr, -bpru,
3542 and -bpra. Now also available in native and exiqgrep formats)
3544 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3545 with very large, slow to parse queues
3547 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3549 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3551 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3552 responses to authentication challenges, though it was showing the
3553 challenges; (ii) I've removed the CR characters from the debug output for
3556 PH/58 Allow for the insertion of a newline as well as a space when a string
3557 is turned into more than one encoded-word during RFC 2047 encoding. The
3558 Sieve code now uses this.
3560 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3561 data_4xx, lost_connection, tls_required.
3563 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3564 message was being sent as an SMTP response.
3566 PH/61 Add -l and -k options to exicyclog.
3568 PH/62 When verifying, if an address was redirected to one new address, so that
3569 verification continued, and the new address failed or deferred after
3570 having set something in $address_data, the value of $address_data was not
3571 passed back to the ACL. This was different to the case when no
3572 redirection occurred. The value is now passed back in both cases.
3574 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3575 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3576 use login_cap.h, so on its own it isn't the distinguishing feature. The
3577 new name refers directly to the setclassresources() function.
3579 PH/65 Added configuration files for NetBSD3.
3581 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3583 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3584 is preferred over IPv4.
3586 PH/68 The bounce_return_message and bounce_return_body options were not being
3587 honoured for bounces generated during the reception of non-SMTP messages.
3588 In particular, this applied to messages rejected by the ACL. This bug has
3589 been fixed. However, if bounce_return_message is true and bounce_return_
3590 body is false, the headers that are returned for a non-SMTP message
3591 include only those that have been read before the error was detected.
3592 (In the case of an ACL rejection, they have all been read.)
3594 PH/69 The HTML version of the specification is now built in a directory called
3595 spec_html instead of spec.html, because the latter looks like a path with
3596 a MIME-type, and this confuses some software.
3598 PH/70 Catch two compiler warnings in sieve.c.
3600 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3601 function verify_get_ident() calls ip_connect() to connect a socket, but
3602 if the "connect()" function timed out, ip_connect() used to close the
3603 socket. However, verify_get_ident() also closes the socket later, and in
3604 between Exim writes to the log, which may get opened at this point. When
3605 the socket was closed in ip_connect(), the log could get the same file
3606 descriptor number as the socket. This naturally causes chaos. The fix is
3607 not to close the socket in ip_connect(); the socket should be closed by
3608 the function that creates it. There was only one place in the code where
3609 this was missing, in the iplookup router, which I don't think anybody now
3610 uses, but I've fixed it anyway.
3612 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3613 well as to direct DNS lookups. Otherwise the handling of names in host
3614 lists is inconsistent and therefore confusing.
3620 PH/01 Two changes to the default runtime configuration:
3622 (1) Move the checks for relay_from_hosts and authenticated clients from
3623 after to before the (commented out) DNS black list checks.
3625 (2) Add control=submission to the relay_from_hosts and authenticated
3626 clients checks, on the grounds that messages accepted by these
3627 statements are most likely to be submissions.
3629 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3631 (1) Generate an error if the third argument for the ${prvs expansion is
3634 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3637 (3) Reset the variables that are obtained from the first argument of
3638 ${prvscheck and used in the second argument before leaving the code,
3639 because their memory is reclaimed, so using them afterwards may do
3642 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3643 one (it's much easier than Tom thought :-).
3645 (5) Because of (4), we can now allow for the use of $prvscheck_result
3646 inside the third argument.
3648 PH/03 For some reason, the default setting of PATH when running a command from
3649 a pipe transport was just "/usr/bin". I have changed it to
3652 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3653 anything to be listed in the output from -bV.
3655 PH/05 When a filter generated an autoreply, the entire To: header line was
3656 quoted in the delivery log line, like this:
3658 => >A.N.Other <ano@some.domain> <original@ddress> ...
3660 This has been changed so that it extracts the operative address. There
3661 may be more than one such address. If so, they are comma-separated, like
3664 => >ano@some.domain,ona@other.domain <original@ddress> ...
3666 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3667 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3668 not being looked up in the rDNS to get a host name, Exim was showing the
3669 IP address twice in Received: lines, even though the IP addresses were
3670 identical. For example:
3672 Received: from [1.2.3.4] (helo=[1.2.3.4])
3674 However, if the real host name was known, it was omitting the HELO data
3675 if it matched the actual IP address. This has been tidied up so that it
3676 doesn't show the same IP address twice.
3678 PH/07 When both +timestamp and +memory debugging was on, the value given by
3679 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3680 called by the debug printing, thereby overwriting the timestamp buffer.
3681 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3683 PH/08 When the original message was included in an autoreply transport, it
3684 always said "this is a copy of the message, including all the headers",
3685 even if body_only or headers_only was set. It now gives an appropriate
3688 PH/09 Applied a patch from the Sieve maintainer which:
3690 o fixes some comments
3691 o adds the (disabled) notify extension core
3692 o adds some debug output for the result of if/elsif tests
3693 o points to the current vacation draft in the documentation
3694 and documents the missing references header update
3698 o fixes a bug in processing the envelope test (when testing
3699 multiple envelope elements, the last element determined the
3702 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3703 Electronic Mail") by including:
3705 Auto-submitted: auto-generated
3707 in the messages that it generates (bounce messages and others, such as
3708 warnings). In the case of bounce messages for non-SMTP messages, there was
3709 also a typo: it was using "Auto_submitted" (underscore instead of
3710 hyphen). Since every message generated by Exim is necessarily in response
3711 to another message, thes have all been changed to:
3713 Auto-Submitted: auto-replied
3715 in accordance with these statements in the RFC:
3717 The auto-replied keyword:
3719 - SHOULD be used on messages sent in direct response to another
3720 message by an automatic process,
3722 - MUST NOT be used on manually-generated messages,
3724 - MAY be used on Delivery Status Notifications (DSNs) and Message
3725 Disposition Notifications (MDNs),
3727 - MUST NOT be used on messages generated by automatic or periodic
3728 processes, except for messages which are automatic responses to
3731 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3732 to the default Received: header definition.
3734 PH/12 Added log selector acl_warn_skipped (default on).
3736 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3737 variables because (a) they are in the wrong storage pool and (b) even if
3738 they were copied, it wouldn't work properly because of the caching.
3740 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3741 checking when decoding. Apparently there are clients that generate
3742 overlong encoded strings. Why am I not surprised?
3744 PH/15 If the first argument of "${if match_address" was not empty, but did not
3745 contain an "@" character, Exim crashed. Now it writes a panic log message
3746 and treats the condition as false.
3748 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3750 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3751 extension "envelope-auth". The code is finished and in agreement with
3752 other implementations, but there is no documentation so far and in fact,
3753 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3754 not changing the active code.
3756 Print executed "if" and "elsif" statements when debugging is used. This
3757 helps a great deal to understand what a filter does.
3759 Document more things not specified clearly in RFC3028. I had all this
3760 sorted out, when out of a sudden new issues came to my mind. Oops."
3762 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3765 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3766 canonical form (as documented). However, after a host name lookup from
3767 the IP address, check_host() was doing a simple string comparison with
3768 addresses acquired from the DNS when checking that the found name did
3769 have the original IP as one of its addresses. Since any found IPv6
3770 addresses are likely to be in abbreviated form, the comparison could
3771 fail. Luckily, there already exists a function for doing the comparison
3772 by converting both addresses to binary, so now that is used instead of
3773 the text comparison.
3775 PH/20 There was another similar case to PH/19, when a complete host name was
3776 given in a host list; looking up its IP address could give an abbreviated
3777 form, whereas the current host's name might or might not be abbreviated.
3778 The same fix has been applied.
3784 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3785 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3788 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3789 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3791 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3793 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3794 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3795 Also, for "Auto-Submitted", treat anything other than "no" as
3796 non-personal, in accordance with RFC 3834. (Previously it treated
3797 anything starting "auto-" as non-personal.)
3799 TF/01 The control=submission/name=... option had a problem with syntax
3800 errors if the name included a slash character. The /name= option
3801 now slurps the rest of the string, so it can include any characters
3802 but it must come last in the list of options (after /sender_retain
3805 PH/05 Some modifications to the interface to the fake nameserver for the new
3813 TK/01 Added the "success_on_redirect" address verification option. See
3814 NewStuff for rationale and an example.
3816 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3818 PH/02 Patch to exigrep to allow it to work on syslog lines.
3820 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3821 fread() to skip over the body file's header line, because in Cygwin the
3822 header line is locked and is inaccessible.
3824 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3825 co-exist for some time) to make it clear that it is the Exim ID that is
3826 referenced, not the Message-ID: header line.
3828 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3829 string_format() function, because snprintf() does not exist on all
3832 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3833 ${stat: expansion item.
3835 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3836 protocol synchronization error", to keep the pedants happy.
3838 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3839 well as for IRIX systems, when gcc is being used. See the host.c source
3842 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3844 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3847 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3848 transport and to the smtp transport in LMTP mode.
3850 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3852 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3853 run a filter in a subprocess. This could lead to confusion in subsequent
3854 lookups in the parent process. There should also be a search_tidyup() at
3855 the end of the subprocess.
3857 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3858 only if the host matched helo_try_verify_hosts, which caused the
3859 verification to occur when the EHLO/HELO command was issued. The ACL just
3860 tested the remembered result. Now, if a previous verification attempt has
3861 not happened, "verify = helo" does it there and then.
3863 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3865 TK/03 Fix log output including CR from clamd.
3867 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3868 exist provoked a memory error which could cause a segfault.
3870 PH/15 Installed PCRE 6.2
3872 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3874 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3875 of the problem. Specifically, suggested +O2 rather than +O1 for the
3878 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3880 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3881 record, the port was not correctly logged when the outgoing_port log
3882 selector was set (it logged the transort's default port).
3884 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3885 fallback_hosts, and "hosts" in the smtp transport.
3887 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3888 host errors such as "Connection refused".
3890 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3891 authentication with radiusclient 0.4.9:
3893 - Error returned from rc_read_config was caught wrongly
3894 - Username/password not passed on to radius server due to wrong length.
3896 The presumption is that some radiusclient API changes for 4.51/PH/17
3897 were not taken care of correctly. The code is still untested by me (my
3898 Linux distribution still has 0.3.2 of radiusclient), but it was
3899 contributed by a Radius user.
3901 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3902 expanding the "port" option of the smtp transport.
3904 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3905 while reading a MIME header. Thanks to Tom Hughes for a patch.
3907 PH/24 Include config.h inside local_scan.h so that configuration settings are
3910 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3911 that in an ACL for RCPT (for example), you can examine exactly what was
3914 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3915 commands, but it was not correctly comparing the address with the actual
3916 client host address. Thus, it would show the EHLO address in Received:
3917 header lines when this was not necessary.
3919 PH/27 Added the % operator to ${eval:}.
3921 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3922 it should be ignoring failures (because with -C, for example, it has lost
3923 privilege). It wasn't ignoring creation failures other than "already
3926 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3929 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3930 because an input line was too long, either on its own, or by virtue of
3931 too many continuations, the temporary file was not being removed, and the
3932 return code was incorrect.
3934 PH/31 Missing "BOOL" in function definition in filtertest.c.
3936 PH/32 Applied Sieve patches from the maintainer.
3938 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3940 PH/33 Added "verify = not_blind".
3942 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3943 Local/Makefile (with some defaults set). These are used in built scripts
3944 such as exicyclog, but they have never been used in the exim_install
3945 script (though there are many overriding facilities there). I have
3946 arranged that the exim_install script now takes note of these two
3949 PH/35 Installed configuration files for Dragonfly.
3951 PH/36 When a locally submitted message by a trusted user did not contain a
3952 From: header, and the sender address was obtained from -f or from an SMTP
3953 MAIL command, and the trusted user did not use -F to supply a sender
3954 name, $originator_name was incorrectly used when constructing a From:
3955 header. Furthermore, $originator_name was used for submission mode
3956 messages from external hosts without From: headers in a similar way,
3957 which is clearly wrong.
3959 PH/37 Added control=suppress_local_fixups.
3961 PH/38 When log_selector = +received_sender was set, and the addition of the
3962 sender made the log line's construction buffer exactly full, or one byte
3963 less than full, an overflow happened when the terminating "\n" was
3966 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3967 when the result of a list match is failure because a DNS lookup failed.
3969 PH/40 RM_COMMAND is now used in the building process.
3971 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3972 the "build-* directories that it finds.
3974 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3975 address in a domain literal was a prefix of an interface address.
3977 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3978 when verifying a sender address, unless rewrite_headers is false.
3980 PH/44 Wrote a long comment about why errors_to addresses are verified as
3981 recipients, not senders.
3983 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3984 the ratelimit ACL was added.
3986 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3988 PH/46 Added extra information about PostgreSQL errors to the error string.
3990 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3991 avoiding the need to install special zones in a real server. This is
3992 backwards compatible; if it can't find the fake resolver, it drops back.
3993 Thus, both old and new test suites can be run.
3995 TF/02 Added util/ratelimit.pl
3997 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3998 clock is set back in time.
4000 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4001 Candler <B.Candler@pobox.com>.
4003 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4004 OK for addresses that are the result of redirections.
4006 PH/48 A number of further additions for the benefit of the new test suite,
4007 including a fake gethostbyname() that interfaces to the fake DNS resolver
4010 TF/06 The fix for widen_domains has also been applied to qualify_single and
4011 search_parents which are the other dnslookup options that can cause
4014 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4017 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4019 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4020 service" if it encountered an MX record with an empty target hostname.
4021 The message is now "an MX or SRV record indicated no SMTP service".
4023 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4024 if the DNS of the sending site is misconfigured. This is quite a
4025 common situation. This change restores the behaviour of treating a
4026 helo verification defer as a failure.
4028 PH/51 If self=fail was set on a router, the bounce message did not include the
4029 actual error message.
4035 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4037 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4038 command itself ended in error, the underlying message about the transport
4039 filter timeout was being overwritten with the pipe command error. Now the
4040 underlying error message should be appended to the second error message.
4042 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4044 PH/02 Reduce the amount of output that "make" produces by default. Full output
4045 can still be requested.
4047 PH/03 The warning log line about a condition test deferring for a "warn" verb
4048 was being output only once per connection, rather than after each
4049 occurrence (because it was using the same function as for successful
4050 "warn" verbs). This seems wrong, so I have changed it.
4052 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4053 it should not have, which might have caused a crash in the right
4054 circumstances, but probably never did.
4056 PH/04 Installed a modified version of Tony Finch's patch to make submission
4057 mode fix the return path as well as the Sender: header line, and to
4058 add a /name= option so that you can make the user's friendly name appear
4061 TF/03 Added the control = fakedefer ACL modifier.
4063 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4064 Mark Lowes for thorough testing.
4066 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4068 TK/03 Merged latest SRS patch from Miles Wilton.
4070 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4071 with the definition in sysexits.h (which is #included earlier).
4072 Fortunately, Exim does not actually use EX_OK. The code used to try to
4073 preserve the sysexits.h value, by assuming that macro definitions were
4074 scanned for macro replacements. I have been disabused of this notion,
4075 so now the code just undefines EX_OK before #including unistd.h.
4077 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4078 in the smtp transport. When a block could not be written in a single
4079 write() function, the timeout was being re-applied to each part-write.
4080 This seems wrong - if the receiver was accepting one byte at a time it
4081 would take for ever. The timeout is now adjusted when this happens. It
4082 doesn't have to be particularly precise.
4084 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4085 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4087 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4088 without a domain if the check to <postmaster@domain> fails.
4090 SC/01 Eximstats: added -xls and the ability to specify output files
4091 (patch written by Frank Heydlauf).
4093 SC/02 Eximstats: use FileHandles for outputting results.
4095 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4097 SC/04 Eximstats: fixed display of large numbers with -nvr option
4099 SC/05 Eximstats: fixed merging of reports with empty tables.
4101 SC/06 Eximstats: added the -include_original_destination flag
4103 SC/07 Eximstats: removed tabs and trailing whitespace.
4105 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4107 TK/06 MBOX spool code: Add real "From " MBOX separator line
4108 so the .eml file is really in mbox format (even though
4109 most programs do not really care). Patch from Alex Miller.
4111 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4112 The latter is generated from $received_to and is only set if the
4113 message has one envelope recipient. SA can use these headers,
4114 obviously out-of-the-box. Patch from Alex Miller.
4116 PH/08 The ${def test on a variable was returning false if the variable's
4117 value was "0", contrary to what the specification has always said!
4118 The result should be true unless the variable is empty.
4120 PH/09 The syntax error of a character other than { following "${if
4121 def:variable_name" (after optional whitespace) was not being diagnosed.
4122 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4123 accidental colon was present, for example, could give incorrect results.
4125 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4126 result is used (not including appendfile, where other changes are about
4129 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4130 This involved changing a lot of size variables from int to off_t. It
4131 should work with maildirs and everything.
4133 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4134 spamd dying while we are connected to it.
4136 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4139 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4140 function for BDB 4.3.
4142 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4144 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4145 the sources for PCRE have been re-arranged and more files are now
4148 PH/15 The code I had for printing potentially long long variables in PH/11
4149 above was not the best (it lost precision). The length of off_t variables
4150 is now inspected at build time, and an appropriate printing format (%ld
4151 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4152 to be "long long int" or "long int". This is needed for the internal
4153 formatting function string_vformat().
4155 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4156 the configuration file to be ":syslog", then the script "guesses" where
4157 the logs files are, rather than using the compiled in default. In our
4158 case the guess is not the same as the compiled default, so the script
4159 suddenly stopped working when I started to use syslog. The patch checks
4160 to see if log_file_path is "". If so, it attempts to read it from exim
4161 with no configuration file to get the compiled in version, before it
4162 falls back to the previous guessing code."
4164 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4165 implementing BATV in an Exim configuration. See NewStuff for the gory
4168 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4169 Makefile that are specific to HP-UX.
4171 PH/18 If the "use_postmaster" option was set for a recipient callout together
4172 with the "random" option, the postmaster address was used as the MAIL
4173 FROM address for the random test, but not for the subsequent recipient
4174 test. It is now used for both.
4176 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4177 patch removes a few documentation additions to RFC 3028, because the
4178 latest draft now contains them. It adds the new en;ascii-case comparator
4179 and a new error check for 8bit text in MIME parts. Comparator and
4180 require names are now matched exactly. I enabled the subaddress
4181 extension, but it is not well tested yet (read: it works for me)."
4183 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4184 rework some of the code of TK/09 above to avoid the hardwired use of
4185 "%lld" and "long long". Replaced the call to snprintf() with a call to
4188 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4189 records point to non-existent hosts", "retry timeout exceeded", and
4190 "retry time not reached for any host after a long failure period".
4192 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4193 experimental DomainKeys support:
4195 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4196 (2) On an error such as an illegally used "control", the wrong name for
4197 the control was given.
4199 These problems did NOT occur unless DomainKeys support was compiled.
4201 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4203 PH/24 Added ${if match_ip condition.
4205 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4206 databases so that it will be absolutely obvious if a crash occurs in the
4207 DB library. This is a regular occurrence (often caused by mis-matched
4210 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4211 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4212 picked up on a user's system that detects such things. There doesn't seem
4213 to be a gcc warning option for this - only an attribute that has to be
4214 put on the function's prototype. It seems that in Fedora Core 4 they have
4215 set this on a number of new functions. No doubt there will be more in due
4218 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4219 specify a transport. However, if an address that was verified by such a
4220 router was the subject of a callout, Exim crashed because it tried to
4221 read the rcpt_include_affixes from the non-existent transport. Now it
4222 just assumes that the setting of that option is false. This bug was
4223 introduced by 4.51/PH/31.
4225 PH/28 Changed -d+all to exclude +memory, because that information is very
4226 rarely of interest, but it makes the output a lot bigger. People tend to
4227 do -d+all out of habit.
4229 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4230 code in os-type was giving problems when libc.so lives in lib64, like on
4233 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4234 aren't the modern standard, and it seems that some systems' include files
4235 don't always have them. Exim was already checking for some of the newer
4236 ones like T_AAAA, and defining it itself. I've added checks for all the
4237 record types that Exim uses.
4239 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4240 not automatically generating a new one, as it is supposed to. This
4241 prevented TLS from working. If the file did exist, but contained invalid
4242 data, a new version was generated, as expected. It was only the case of a
4243 non-existent file that was broken.
4245 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4246 with a change in libdomainkeys > 0.64.
4248 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4249 from DNS. If the selector record carries the flag, it now has
4250 precedence over the domain-wide flag.
4252 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4254 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4255 the use of a port name that isn't defined in /etc/services) occurred, the
4256 message was deferred as in a normal delivery, and thus remained on the
4257 spool, instead of being failed because of the mua_wrapper setting. This
4258 is now fixed, and I tidied up some of the mua_wrapper messages at the
4261 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4262 the messages in a hash of arrays rather than using individual hashes.
4263 This is a bit cleaner and results in dramatic memory savings, albeit
4264 at a slight CPU cost.
4266 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4267 as requested by Marc Sherman.
4269 SC/10 Eximstats: added histograms for user specified patterns as requested
4272 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4274 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4275 fopen() in the content-scanning modules that did not already have it.
4281 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4282 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4284 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4286 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4288 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4289 to test Sieve filters that use "vacation".
4291 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4292 that changes the way the GnuTLS parameters are stored in the cache file.
4293 The new format can be generated externally. For backward compatibility,
4294 if the data in the cache doesn't make sense, Exim assumes it has read an
4295 old-format file, and it generates new data and writes a new file. This
4296 means that you can't go back to an older release without removing the
4299 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4300 work if there are any delivery delays because "one_time" forces the
4301 parent to be marked "delivered", so its unseen clone is never tried
4302 again. For this reason, Exim now forbids the simultaneous setting of
4305 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4306 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4307 entry if you want to know the details. The fix, however, neglected to
4308 consider the case when local delivery batching is involved. The test for
4309 "previously delivered" was not happening when checking to see if an
4310 address could be batched with a previous (undelivered) one; under
4311 certain circumstances this could lead to multiple deliveries to the same
4314 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4315 in its include files, and this causes problems building Exim.
4317 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4318 header_syntax) but Exim was just ignoring anything given after a slash.
4319 In particular, this caused confusion with an attempt to use "verify =
4320 reverse_host_lookup/defer_ok". An error is now given when options are
4321 supplied for verify items that do not have them. (Maybe reverse_host_
4322 lookup should have a defer_ok option, but that's a different point.)
4324 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4325 defined by RFC 821) to 2048, because there were problems with some AUTH
4326 commands, and RFC 1869 says the size should be increased for extended
4327 SMTP commands that take arguments.
4329 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4332 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4333 "unknown" error; now it says that the functionality isn't in the binary.
4335 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4336 an address' error message when a string expansion fails (syntax or
4337 whatever). Otherwise the password may appear in the log. Following change
4338 PH/42 below, there is no longer a chance of it appearing in a bounce
4341 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4343 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4344 of its addresses was ever tried. (Bugzilla bug #2.)
4346 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4347 the result incorrectly in the debug output. (It correctly added a newline
4348 to what was transported.)
4350 TF/01 Added $received_time.
4352 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4353 commented out examples of how to interface to a virus scanner and to
4354 SpamAssassin. Also added commented examples of av_scanner and
4355 spamd_address settings.
4357 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4358 and controls are allowed in which ACLs. There were a couple of minor
4359 errors. Some of the entries in the conditions table (which is a table of
4360 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4361 negation of where the condition IS allowed.
4363 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4365 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4366 header file does not have a version number, so I've had to invent a new
4367 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4368 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4369 radiusclient), but it was contributed by a Radius user.
4371 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4372 files or directories, for OpenSSL.
4374 PH/19 When an Exim process that is running as root has to create an Exim log
4375 file, it does so in a subprocess that runs as exim:exim so as to get the
4376 ownership right at creation (otherwise, other Exim processes might see
4377 the file with the wrong ownership). There was no test for failure of this
4378 fork() call, which would lead to the process getting stuck as it waited
4379 for a non-existent subprocess. Forks do occasionally fail when resources
4380 run out. I reviewed all the other calls to fork(); they all seem to check
4383 PH/20 When checking for unexpected SMTP input at connect time (before writing
4384 the banner), Exim was not dealing correctly with a non-positive return
4385 from the read() function. If the client had disconnected by this time,
4386 the result was a log entry for a synchronization error with an empty
4387 string after "input=" when read() returned zero. If read() returned -1
4388 (an event I could not check), uninitialized data bytes were printed.
4389 There were reports of junk text (parts of files, etc) appearing after
4392 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4394 PH/22 Added support for macro redefinition, and (re)definition in between
4395 driver and ACL definitions.
4397 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4398 forgetting to use the resulting value; it was using the unexpanded value.
4400 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4401 hadn't been configured. The fix is from Juergen Kreileder, who
4402 understands it better than I do:
4404 "Here's what I see happening with three configured cyrus_sasl
4405 authenticators configured (plain, login, cram-md5):
4407 On startup auth_cyrus_sasl_init() gets called for each of these.
4408 This means three calls to sasl_listmech() without a specified mech_list.
4409 => SASL tests which mechs of all available mechs actually work
4410 => three warnings about OTP not working
4411 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4413 With the patch, sasl_listmech() also gets called three times. But now
4414 SASL's mech_list option is set to the server_mech specified in the the
4415 authenticator. Or in other words, the answer from sasl_listmech()
4416 gets limited to just the mech you're testing for (which is different
4418 => the return list contains just 'plain' or 'login', 'cram-md5' or
4419 nothing depending on the value of ob->server_mech.
4421 I've just tested the patch: Authentication still works fine,
4422 unavailable mechs specified in the exim configuration are still
4423 caught, and the auth.log warnings about OTP are gone."
4425 PH/25 When debugging is enabled, the contents of the command line are added
4426 to the debugging output, even when log_selector=+arguments is not
4429 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4430 answer is "GNU", and only if the return is "GNU/something" is the answer
4433 PH/27 $acl_verify_message is now set immediately after the failure of a
4434 verification in an ACL, and so is available in subsequent modifiers. In
4435 particular, the message can be preserved by coding like this:
4437 warn !verify = sender
4438 set acl_m0 = $acl_verify_message
4440 Previously, $acl_verify_message was set only while expanding "message"
4441 and "log_message" when a very denied access.
4443 PH/28 Modified OS/os.c-Linux with
4445 -#ifndef OS_LOAD_AVERAGE
4446 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4448 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4449 nomenclature these days.)
4451 PH/29 Installed patch from the Sieve maintainer that adds the options
4452 sieve_useraddress and sieve_subaddress to the redirect router.
4454 PH/30 In these circumstances:
4455 . Two addresses routed to the same list of hosts;
4456 . First host does not offer TLS;
4457 . First host accepts first address;
4458 . First host gives temporary error to second address;
4459 . Second host offers TLS and a TLS session is established;
4460 . Second host accepts second address.
4461 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4462 and peerdn, if requested) that were in fact used only for the second
4465 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4466 attention to any local part prefix or suffix that was matched by the
4467 router that accepted the address. It now behaves in the same way as it
4468 does for delivery: the affixes are removed from the local part unless
4469 rcpt_include_affixes is set on the transport.
4471 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4472 timeout during the DATA phase of an incoming message.
4474 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4475 applied a patch sanctioned by the Sieve maintainer.
4477 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4478 the uid or gid is negative. A case of a negative gid caused this to be
4479 noticed. The fix allows for either to be negative.
4481 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4482 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4485 PH/36 The change PH/12 above was broken. Fixed it.
4487 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4488 the grounds that routing the same address twice would always produce the
4489 same answer. This might have been true once, but it is certainly no
4490 longer true now. Routing a child address may depend on the previous
4491 routing that produced that child. Some complicated redirection strategies
4492 went wrong when messages had multiple recipients, and made Exim's
4493 behaviour dependent on the order in which the addresses were given.
4495 I have moved the duplicate checking until after the routing is complete.
4496 Exim scans the addresses that are assigned to local and remote
4497 transports, and removes any duplicates. This means that more work will be
4498 done, as duplicates will always all be routed, but duplicates are
4499 presumably rare, so I don't expect this is of any significance.
4501 For deliveries to pipes, files, and autoreplies, the duplicate checking
4502 still happens during the routing process, since they are not going to be
4505 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4506 It corrects a timeout issue with spamd. This is Ian's comment: "The
4507 background is that sometimes spamd either never reads data from a
4508 connection it has accepted, or it never writes response data. The exiscan
4509 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4510 blindly assumes that writes won't block so it may never time out."
4512 PH/39 Allow G after quota size as well as K and M.
4514 PH/40 The value set for $authenticated_id in an authenticator may not contain
4515 binary zeroes or newlines because the value is written to log lines and
4516 to spool files. There was no check on this. Now the value is run through
4517 the string_printing() function so that such characters are converted to
4518 printable escape sequences.
4520 PH/41 $message_linecount is a new variable that contains the total number of
4521 lines in the message. Compare $body_linecount, which is the count for the
4524 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4525 bounce and delay warning messages, except in certain special cases, which
4528 (a) An SMTP error message from a remote host;
4529 (b) A message specified in a :fail: redirection;
4530 (c) A message specified in a "fail" command in a system filter;
4531 (d) A message specified in a FAIL return from the queryprogram router;
4532 (e) A message specified by the cannot_route_message router option.
4534 In these cases only, Exim does include the error details in bounce and
4535 warning messages. There are also a few cases where bland messages such
4536 as "unrouteable address" or "local delivery error" are given.
4538 PH/43 $value is now also set for the "else" part of a ${run expansion.
4540 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4541 being worked on, but at least Exim now implements the latest version to
4544 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4545 process to complete was treated as a delivery failure, a timeout while
4546 writing the message to the pipe was logged, but erroneously treated as a
4547 successful delivery. Such timeouts include transport filter timeouts. For
4548 consistency with the overall process timeout, these timeouts are now
4549 treated as errors, giving rise to delivery failures by default. However,
4550 there is now a new Boolean option for the pipe transport called
4551 timeout_defer, which, if set TRUE, converts the failures into defers for
4552 both kinds of timeout. A transport filter timeout is now identified in
4555 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4556 systems where "make" and "gmake" are different, calling "gmake" at top
4557 level broke things. I've arranged for the value of $(MAKE) to be passed
4558 from the Makefile to this script so that it can call the same version of
4562 A note about Exim versions 4.44 and 4.50
4563 ----------------------------------------
4565 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4566 changes of various kinds. As a consequence, a big documentation update was
4567 needed. This delayed the release for rather longer than seemed good, especially
4568 in the light of a couple of (minor) security issues. Therefore, the changes
4569 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4570 release. So 4.44 and 4.50 are in effect two different branches that both start
4573 I have left the 4.50 change log unchanged; it contains all the changes since
4574 4.43. The change log for 4.44 is below; many of its items are identical to
4575 those for 4.50. This seems to be the most sensible way to preserve the
4576 historical information.
4582 1. Minor wording change to the doc/README.SIEVE file.
4584 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4585 computation of the current number of files was incorrect.
4587 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4588 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4591 4. Give more explanation in the error message when the command for a transport
4592 filter fails to execute.
4594 5. There are several places where Exim runs a non-Exim command in a
4595 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4596 was being done only for the command run by the queryprogram router. It is
4597 now done for all such subprocesses. The other cases are: ${run, transport
4598 filters, and the commands run by the lmtp and pipe transports.
4600 6. Added CONFIGURE_GROUP build-time option.
4602 7. Some older OS have a limit of 256 on the maximum number of file
4603 descriptors. Exim was using setrlimit() to set 1000 as a large value
4604 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4605 systems. I've change it so that if it can't get 1000, it tries for 256.
4607 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4608 was an oversight, and furthermore, ever since the addition of extra
4609 controls (e.g. 4.43/32), the checks on when to allow different forms of
4610 "control" were broken. There should now be diagnostics for all cases when a
4611 control that does not make sense is encountered.
4613 9. Added the /retain_sender option to "control=submission".
4615 10. $recipients is now available in the predata ACL (oversight).
4617 11. Tidy the search cache before the fork to do a delivery from a message
4618 received from the command line. Otherwise the child will trigger a lookup
4619 failure and thereby defer the delivery if it tries to use (for example) a
4620 cached ldap connection that the parent has called unbind on.
4622 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4623 of $address_data from the recipient verification was clobbered by the
4624 sender verification.
4626 13. The value of address_data from a sender verification is now available in
4627 $sender_address_data in subsequent conditions in the ACL statement.
4629 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4631 15. Added a new option "connect=<time>" to callout options, to set a different
4634 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4635 was its contents. (It was OK if the option was not defined at all.)
4637 17. A "Completed" log line is now written for messages that are removed from
4638 the spool by the -Mrm option.
4640 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4641 information about exactly what failed.
4643 19. Added -dd to debug only the daemon process.
4645 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4646 handles timeouts, both on the server side and network timeouts. Renamed the
4647 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4649 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4650 It is now set to "smtps".
4652 22. $host_address is now set to the target address during the checking of
4653 ignore_target_hosts.
4655 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4656 being passed; this would have caused $sender_host_name to have been used if
4657 matching the list had actually called for a host name (not very likely,
4658 since this list is usually IP addresses). A host name is now passed as
4661 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4662 code that specifies a non-restarting handler (typically sigaction() in
4663 modern systems) in an attempt to fix a rare and obscure crash bug.
4665 25. Narrowed the window for a race in the daemon that could cause it to ignore
4666 SIGCHLD signals. This is not a major problem, because they are used only to
4667 wake it up if nothing else does.
4669 26. A malformed maildirsize file could cause Exim to calculate negative values
4670 for the mailbox size or file count. Odd effects could occur as a result.
4671 The maildirsize information is now recalculated if the size or filecount
4674 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4675 support for a long time. Removed HAVE_SYS_VFS_H.
4677 28. Installed the latest version of exipick from John Jetmore.
4679 29. In an address list, if the pattern was not a regular expression, an empty
4680 subject address (from a bounce message) matched only if the pattern was an
4681 empty string. Non-empty patterns were not even tested. This was the wrong
4682 because it is perfectly reasonable to use an empty address as part of a
4683 database query. An empty address is now tested by patterns that are
4684 lookups. However, all the other forms of pattern expect the subject to
4685 contain a local part and a domain, and therefore, for them, an empty
4686 address still always fails if the pattern is not itself empty.
4688 30. Exim went into a mad DNS loop when attempting to do a callout where the
4689 host was specified on an smtp transport, and looking it up yielded more
4690 than one IP address.
4692 31. Re-factored the code for checking spool and log partition space into a
4693 function that finds that data and another that does the check. The former
4694 is then used to implement four new variables: $spool_space, $log_space,
4695 $spool_inodes, and $log_inodes.
4697 32. The RFC2047 encoding function was originally intended for short strings
4698 such as real names; it was not keeping to the 75-character limit for
4699 encoded words that the RFC imposes. It now respects the limit, and
4700 generates multiple encoded words if necessary. To be on the safe side, I
4701 have increased the buffer size for the ${rfc2047: expansion operator from
4704 33. It is now permitted to omit both strings after an "if" condition; if the
4705 condition is true, the result is "true". As before, when the second string
4706 is omitted, a false condition yields an empty string. This makes it less
4707 cumbersome to write custom ACL and router conditions.
4709 34. Failure to deliver a bounce message always caused it to be frozen, even if
4710 there was an errors_to setting on the router. The errors_to setting is now
4713 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4714 canonical form (fully expanded) before being placed in
4715 $sender_host_address.
4717 36. The table in the code that translates DNS record types into text (T_A to
4718 "A" for instance) was missing entries for NS and CNAME. It is just possible
4719 that this could have caused confusion if both these types were looked up
4720 for the same domain, because the text type is used as part of Exim's
4721 per-process caching. But the chance of anyone hitting this buglet seems
4724 37. The dnsdb lookup has been extended in a number of ways.
4726 (1) There is a new type, "zns", which walks up the domain tree until it
4727 finds some nameserver records. It should be used with care.
4729 (2) There is a new type, "mxh", which is like "mx" except that it returns
4730 just the host names, not the priorities.
4732 (3) It is now possible to give a list of domains (or IP addresses) to be
4733 looked up. The behaviour when one of the lookups defers can be
4734 controlled by a keyword.
4736 (4) It is now possible to specify the separator character for use when
4737 multiple records are returned.
4739 38. The dnslists ACL condition has been extended: it is now possible to supply
4740 a list of IP addresses and/or domains to be looked up in a particular DNS
4743 39. Added log_selector=+queue_time_overall.
4745 40. When running the queue in the test harness, wait just a tad after forking a
4746 delivery process, to get repeatability of debugging output.
4748 41. Include certificate and key file names in error message when GnuTLS fails
4749 to set them up, because the GnuTLS error message doesn't include the name
4750 of the failing file when there is a problem reading it.
4752 42. Allow both -bf and -bF in the same test run.
4754 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4756 44. The "Exiscan patch" is now merged into the mainline Exim source.
4758 45. Sometimes the final signoff response after QUIT could fail to get
4759 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4760 < 0 before doing a fflush(). This bug looks as though it goes back to the
4761 introduction of TLS in release 3.20, but "sometimes" must have been rare
4762 because the tests only now provoked it.
4764 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4765 (this can affect the format of dates).
4767 47. exim_tidydb, when checking for the continued existence of a message for
4768 which it has found a message-specific retry record, was not finding
4769 messages that were in split spool directories. Consequently, it was
4770 deleting retry records that should have stayed in existence.
4772 48. Steve fixed some bugs in eximstats.
4774 49. The SPA authentication driver was not abandoning authentication and moving
4775 on to the next authenticator when an expansion was forced to fail,
4776 contradicting the general specification for all authenticators. Instead it
4777 was generating a temporary error. It now behaves as specified.
4779 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4780 (the order specifies the preference for clients). The order is now AES256,
4781 AES128, 3DES, ARCFOUR128.
4783 51. Small patch to Sieve code - explicitly set From: when generating an
4786 52. Exim crashed if a remote delivery caused a very long error message to be
4787 recorded - for instance if somebody sent an entire SpamAssassin report back
4788 as a large number of 550 error lines. This bug was coincidentally fixed by
4789 increasing the size of one of Exim's internal buffers (big_buffer) that
4790 happened as part of the Exiscan merge. However, to be on the safe side, I
4791 have made the code more robust (and fixed the comments that describe what
4794 53. Now that there can be additional text after "Completed" in log lines (if
4795 the queue_time_overall log selector is set), a one-byte patch to exigrep
4796 was needed to allow it to recognize "Completed" as not the last thing in
4799 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4800 patch that reportedly fixes this has been added. I am not expert enough to
4801 create a test for it. This is what the patch creator wrote:
4803 "I found a little strange behaviour of ldap code when working with
4804 Windows 2003 AD Domain, where users was placed in more than one
4805 Organization Units. When I tried to give exim partial DN, the exit code
4806 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4807 But simultaneously result of request was absolutely normal ldap result,
4808 so I produce this patch..."
4810 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4811 REFERENCE, so I have modified the code to exclude the patch when that macro
4814 55. Some experimental protocols are using DNS PTR records for new purposes. The
4815 keys for these records are domain names, not reversed IP addresses. The
4816 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4817 leaves it alone. Component reversal etc. now happens only for IP addresses.
4820 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4822 57. Double the size of the debug message buffer (to 2048) so that more of very
4823 long debug lines gets shown.
4825 58. The exicyclog utility now does better if the number of log files to keep
4826 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4828 59. Two changes related to the smtp_active_hostname option:
4830 (1) $smtp_active_hostname is now available as a variable.
4831 (2) The default for smtp_banner uses $smtp_active_hostname instead
4832 of $primary_hostname.
4834 60. The host_aton() function is supposed to be passed a string that is known
4835 to be a valid IP address. However, in the case of IPv6 addresses, it was
4836 not checking this. This is a hostage to fortune. Exim now panics and dies
4837 if the condition is not met. A case was found where this could be provoked
4838 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4839 components; fortuitously, this particular loophole had already been fixed
4840 by change 4.50/55 above.
4842 If there are any other similar loopholes, the new check in host_aton()
4843 itself should stop them being exploited. The report I received stated that
4844 data on the command line could provoke the exploit when Exim was running as
4845 exim, but did not say which command line option was involved. All I could
4846 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4847 running as the user.
4850 61. There was a buffer overflow vulnerability in the SPA authentication code
4851 (which came originally from the Samba project). I have added a test to the
4852 spa_base64_to_bits() function which I hope fixes it.
4855 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4856 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4858 63. The daemon start-up calls getloadavg() while still root for those OS that
4859 need the first call to be done as root, but it missed one case: when
4860 deliver_queue_load_max is set with deliver_drop_privilege. This is
4861 necessary for the benefit of the queue runner, because there is no re-exec
4862 when deliver_drop_privilege is set.
4864 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4865 This has been fixed.
4867 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4868 was in use, was not putting the data itself into the right store pool;
4869 consequently, it could be overwritten for a subsequent message in the same
4870 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4873 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4875 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4876 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4877 the function were treating the return as a boolean value, which happened to
4878 work because 0=false and not-0=true, but is not correct code.
4880 68. The host_aton() function was not handling scoped IPv6 addresses (those
4881 with, for example, "%eth0" on the end) correctly.
4883 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4884 negated items (that is, ~something) in unsigned ints. Some compilers
4885 apparently mutter when there is no cast.
4887 70. If an address verification called from an ACL failed, and did not produce a
4888 user-specific message (i.e. there was only a "system" message), nothing was
4889 put in $acl_verify_message. In this situation, it now puts the system
4892 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4893 session; change 4.31/43 added the unwanted input to the log line - except
4894 that it did not do this in the start of session case. It now does.
4896 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4897 This is wrong and can cause the other end to generate a synchronization
4898 error if it is another Exim or anything else that does the synchronization
4899 check. A QUIT command is no longer sent after a timeout.
4901 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4902 during host lookups.
4904 74. The defer_ok option of callout verification was not working if it was used
4905 when verifying addresses in header lines, that is, for this case:
4907 verify = header_sender/callout=defer_ok
4909 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4910 those file descriptors could be used for SMTP connections. If anything
4911 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4912 could be sent to the SMTP client, causing chaos. The daemon now opens
4913 stdin, stdout, and stderr to /dev/null when it puts itself into the
4916 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4917 log by default. The user can override this with suitable Perl magic.
4919 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4920 the log message when discard triggers, was not working for the DATA ACL or
4921 for the non-SMTP ACL.
4923 78. Error message wording change in sieve.c.
4925 79. If smtp_accept_max_per_host was set, the number of connections could be
4926 restricted to fewer than expected, because the daemon was trying to set up
4927 a new connection before checking whether the processes handling previous
4928 connections had finished. The check for completed processes is now done
4929 earlier. On busy systems, this bug wouldn't be noticed because something
4930 else would have woken the daemon, and it would have reaped the completed
4933 80. If a message was submitted locally by a user whose login name contained one
4934 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4935 It caused a spool format error. I have fixed the spool reading code. A
4936 related problem was that the "from" clause in the Received: line became
4937 illegal because of the space(s). It is now covered by ${quote_local_part.
4939 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4942 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4944 83. Patch from Sieve maintainer to fix unterminated string problem in
4945 "vacation" handling.
4947 84. Some minor changes to the Linux configuration files to help with other
4948 OS variants using glibc.
4950 85. One more patch for Sieve to update vacation handling to latest spec.
4953 ----------------------------------------------------
4954 See the note above about the 4.44 and 4.50 releases.
4955 ----------------------------------------------------
4961 1. Change 4.43/35 introduced a bug that caused file counts to be
4962 incorrectly computed when quota_filecount was set in an appendfile
4965 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4966 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4969 3. Give more explanation in the error message when the command for a transport
4970 filter fails to execute.
4972 4. There are several places where Exim runs a non-Exim command in a
4973 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4974 was being done only for the command run by the queryprogram router. It is
4975 now done for all such subprocesses. The other cases are: ${run, transport
4976 filters, and the commands run by the lmtp and pipe transports.
4978 5. Some older OS have a limit of 256 on the maximum number of file
4979 descriptors. Exim was using setrlimit() to set 1000 as a large value
4980 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4981 systems. I've change it so that if it can't get 1000, it tries for 256.
4983 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4984 was an oversight, and furthermore, ever since the addition of extra
4985 controls (e.g. 4.43/32), the checks on when to allow different forms of
4986 "control" were broken. There should now be diagnostics for all cases when a
4987 control that does not make sense is encountered.
4989 7. $recipients is now available in the predata ACL (oversight).
4991 8. Tidy the search cache before the fork to do a delivery from a message
4992 received from the command line. Otherwise the child will trigger a lookup
4993 failure and thereby defer the delivery if it tries to use (for example) a
4994 cached ldap connection that the parent has called unbind on.
4996 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4997 of $address_data from the recipient verification was clobbered by the
4998 sender verification.
5000 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5001 was its contents. (It was OK if the option was not defined at all.)
5003 11. A "Completed" log line is now written for messages that are removed from
5004 the spool by the -Mrm option.
5006 12. $host_address is now set to the target address during the checking of
5007 ignore_target_hosts.
5009 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5010 being passed; this would have caused $sender_host_name to have been used if
5011 matching the list had actually called for a host name (not very likely,
5012 since this list is usually IP addresses). A host name is now passed as
5015 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5016 code that specifies a non-restarting handler (typically sigaction() in
5017 modern systems) in an attempt to fix a rare and obscure crash bug.
5019 15. Narrowed the window for a race in the daemon that could cause it to ignore
5020 SIGCHLD signals. This is not a major problem, because they are used only to
5021 wake it up if nothing else does.
5023 16. A malformed maildirsize file could cause Exim to calculate negative values
5024 for the mailbox size or file count. Odd effects could occur as a result.
5025 The maildirsize information is now recalculated if the size or filecount
5028 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5029 support for a long time. Removed HAVE_SYS_VFS_H.
5031 18. Updated exipick to current release from John Jetmore.
5033 19. Allow an empty sender to be matched against a lookup in an address list.
5034 Previously the only cases considered were a regular expression, or an
5037 20. Exim went into a mad DNS lookup loop when doing a callout where the
5038 host was specified on the transport, if the DNS lookup yielded more than
5041 21. The RFC2047 encoding function was originally intended for short strings
5042 such as real names; it was not keeping to the 75-character limit for
5043 encoded words that the RFC imposes. It now respects the limit, and
5044 generates multiple encoded words if necessary. To be on the safe side, I
5045 have increased the buffer size for the ${rfc2047: expansion operator from
5048 22. Failure to deliver a bounce message always caused it to be frozen, even if
5049 there was an errors_to setting on the router. The errors_to setting is now
5052 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5053 canonical form (fully expanded) before being placed in
5054 $sender_host_address.
5056 24. Updated eximstats to version 1.33
5058 25. Include certificate and key file names in error message when GnuTLS fails
5059 to set them up, because the GnuTLS error message doesn't include the name
5060 of the failing file when there is a problem reading it.
5062 26. Expand error message when OpenSSL has problems setting up cert/key files.
5065 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5066 (this can affect the format of dates).
5068 28. exim_tidydb, when checking for the continued existence of a message for
5069 which it has found a message-specific retry record, was not finding
5070 messages that were in split spool directories. Consequently, it was
5071 deleting retry records that should have stayed in existence.
5073 29. eximstats updated to version 1.35
5074 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5075 1.35 - bugfix such that pie charts by volume are generated correctly
5077 30. The SPA authentication driver was not abandoning authentication and moving
5078 on to the next authenticator when an expansion was forced to fail,
5079 contradicting the general specification for all authenticators. Instead it
5080 was generating a temporary error. It now behaves as specified.
5082 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5083 (the order specifies the preference for clients). The order is now AES256,
5084 AES128, 3DES, ARCFOUR128.
5086 31. Small patch to Sieve code - explicitly set From: when generating an
5089 32. Exim crashed if a remote delivery caused a very long error message to be
5090 recorded - for instance if somebody sent an entire SpamAssassin report back
5091 as a large number of 550 error lines. This bug was coincidentally fixed by
5092 increasing the size of one of Exim's internal buffers (big_buffer) that
5093 happened as part of the Exiscan merge. However, to be on the safe side, I
5094 have made the code more robust (and fixed the comments that describe what
5097 33. Some experimental protocols are using DNS PTR records for new purposes. The
5098 keys for these records are domain names, not reversed IP addresses. The
5099 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5100 leaves it alone. Component reversal etc. now happens only for IP addresses.
5103 34. The host_aton() function is supposed to be passed a string that is known
5104 to be a valid IP address. However, in the case of IPv6 addresses, it was
5105 not checking this. This is a hostage to fortune. Exim now panics and dies
5106 if the condition is not met. A case was found where this could be provoked
5107 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5108 components; fortuitously, this particular loophole had already been fixed
5109 by change 4.50/55 or 4.44/33 above.
5111 If there are any other similar loopholes, the new check in host_aton()
5112 itself should stop them being exploited. The report I received stated that
5113 data on the command line could provoke the exploit when Exim was running as
5114 exim, but did not say which command line option was involved. All I could
5115 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5116 running as the user.
5119 35. There was a buffer overflow vulnerability in the SPA authentication code
5120 (which came originally from the Samba project). I have added a test to the
5121 spa_base64_to_bits() function which I hope fixes it.
5124 36. The daemon start-up calls getloadavg() while still root for those OS that
5125 need the first call to be done as root, but it missed one case: when
5126 deliver_queue_load_max is set with deliver_drop_privilege. This is
5127 necessary for the benefit of the queue runner, because there is no re-exec
5128 when deliver_drop_privilege is set.
5130 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5131 was in use, was not putting the data itself into the right store pool;
5132 consequently, it could be overwritten for a subsequent message in the same
5133 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5136 38. Sometimes the final signoff response after QUIT could fail to get
5137 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5138 < 0 before doing a fflush(). This bug looks as though it goes back to the
5139 introduction of TLS in release 3.20, but "sometimes" must have been rare
5140 because the tests only now provoked it.
5146 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5147 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5148 Now it returns an integer. A number of calls were still expecting a T/F
5149 return. Fortuitously, in all cases, the tests worked in OK situations,
5150 which is the norm. However, things would have gone wrong on any write
5151 failures on the smtp file descriptor. This function is used when sending
5152 messages over SMTP and also when doing verify callouts.
5154 2. When Exim is called to do synchronous delivery of a locally submitted
5155 message (the -odf or -odi options), it no longer closes stderr before doing
5158 3. Implemented the mua_wrapper option.
5160 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5162 5. Implemented the functions header_remove(), header_testname(),
5163 header_add_at_position(), and receive_remove_recipient(), and exported them
5166 6. If an ACL "warn" statement specified the addition of headers, Exim already
5167 inserted X-ACL-Warn: at the start if there was no header name. However, it
5168 was not making this test for the second and subsequent header lines if
5169 there were newlines in the string. This meant that an invalid header could
5170 be inserted if Exim was badly configured.
5172 7. Allow an ACL "warn" statement to add header lines at the start or after all
5173 the Received: headers, as well as at the end.
5175 8. Added the rcpt_4xx retry error code.
5177 9. Added postmaster_mailfrom=xxx to callout verification option.
5179 10. Added mailfrom=xxxx to the callout verification option, for verify=
5182 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5183 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5185 12. Inserted some casts to stop certain compilers warning when using pointer
5186 differences as field lengths or precisions in printf-type calls (mostly
5187 affecting debugging statements).
5189 13. Added optional readline() support for -be (dynamically loaded).
5191 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5192 same clock tick as a message's arrival, so that its received time was the
5193 same as the "first fail" time on the retry record, and that message
5194 remained on the queue past the ultimate address timeout, every queue runner
5195 would try a delivery (because it was past the ultimate address timeout) but
5196 after another failure, the ultimate address timeout, which should have then
5197 bounced the address, did not kick in. This was a "< instead of <=" error;
5198 in most cases the first failure would have been in the next clock tick
5199 after the received time, and all would be well.
5201 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5202 being recognized when the domain list was tested by the match_domain
5203 condition in an expansion string.
5205 16. Added the ${str2b64: operator.
5207 17. Exim was always calling setrlimit() to set a large limit for the number of
5208 processes, without checking whether the existing limit was already
5209 adequate. (It did check for the limit on file descriptors.) Furthermore,
5210 errors from getrlimit() and setrlimit() were being ignored. Now they are
5211 logged to the main and panic logs, but Exim does carry on, to try to do its
5212 job under whatever limits there are.
5214 18. Imported PCRE 5.0.
5216 19. Trivial typo in log message " temporarily refused connection" (the leading
5219 20. If the log selector return_path_on_delivery was set and an address was
5220 redirected to /dev/null, the delivery process crashed because it assumed
5221 that a return path would always be set for a "successful" delivery. In this
5222 case, the whole delivery is bypassed as an optimization, and therefore no
5225 21. Internal re-arrangement: the function for sending a challenge and reading
5226 a response while authentication was assuming a zero-terminated challenge
5227 string. It's now changed to take a pointer and a length, to allow for
5228 binary data in such strings.
5230 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5232 23. Exim was not respecting finduser_retries when seeking the login of the
5233 uid under which it was called; it was always trying 10 times. (The default
5234 setting of finduser_retries is zero.) Also, it was sleeping after the final
5235 failure, which is pointless.
5237 24. Implemented tls_on_connect_ports.
5239 25. Implemented acl_smtp_predata.
5241 26. If the domain in control=submission is set empty, Exim assumes that the
5242 authenticated id is a complete email address when it generates From: or
5243 Sender: header lines.
5245 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5246 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5247 chown and chgrp in /bin and hostname in /usr/bin.
5249 28. Exim was keeping the "process log" file open after each use, just as it
5250 does for the main log. This opens the possibility of it remaining open for
5251 long periods when the USR1 signal hits a daemon. Occasional processlog
5252 errors were reported, that could have been caused by this. Anyway, it seems
5253 much more sensible not to leave this file open at all, so that is what now
5256 29. The long-running daemon process does not normally write to the log once it
5257 has entered its main loop, and it closes the log before doing so. This is
5258 so that log files can straightforwardly be renamed and moved. However,
5259 there are a couple of unusual error situations where the daemon does write
5260 log entries, and I had neglected to close the log afterwards.
5262 30. The text of an SMTP error response that was received during a remote
5263 delivery was being truncated at 512 bytes. This is too short for some of
5264 the long messages that one sometimes sees. I've increased the limit to
5267 31. It is now possible to make retry rules that apply only when a message has a
5268 specific sender, in particular, an empty sender.
5270 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5271 it possible to be selective about when SMTP synchronization is enforced.
5273 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5275 32. Implemented hosts_connection_nolog.
5277 33. Added an ACL for QUIT.
5279 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5282 35. Added mailbox_size and mailbox_filecount to appendfile.
5284 36. Added control = no_multiline_responses to ACLs.
5286 37. There was a bug in the logic of the code that waits for the clock to tick
5287 in the case where the clock went backwards by a substantial amount such
5288 that the microsecond fraction of "now" was more than the microsecond
5289 fraction of "then" (but the whole seconds number was less).
5291 38. Added support for the libradius Radius client library this is found on
5292 FreeBSD (previously only the radiusclient library was supported).
5298 1. When certain lookups returned multiple values in the form name=value, the
5299 quoting of the values was not always being done properly. Specifically:
5300 (a) If the value started with a double quote, but contained no whitespace,
5302 (b) If the value contained whitespace other than a space character (i.e.
5303 tabs or newlines or carriage returns) it was not quoted.
5304 This fix has been applied to the mysql and pgsql lookups by writing a
5305 separate quoting function and calling it from the lookup code. The fix
5306 should probably also be applied to nisplus, ibase and oracle lookups, but
5307 since I cannot test any of those, I have not disturbed their existing code.
5309 2. A hit in the callout cache for a specific address caused a log line with no
5310 reason for rejecting RCPT. Now it says "Previous (cached) callout
5311 verification failure".
5313 3. There was an off-by-one bug in the queryprogram router. An over-long
5314 return line was truncated at 256 instead of 255 characters, thereby
5315 overflowing its buffer with the terminating zero. As well as fixing this, I
5316 have increased the buffer size to 1024 (and made a note to document this).
5318 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5319 when Exim is waiting for an SMTP response from a remote server, Exim
5320 restarts its select() call on the socket, thereby resetting its timeout.
5321 This is not a problem when such interrupts are rare. Somebody set up a cron
5322 job to run exiwhat every 2 minutes, which is less than the normal select()
5323 timeout (5 or 10 minutes). This meant that the select() timeout never
5324 kicked in because it was always reset. I have fixed this by comparing the
5325 time when an interrupt arrives with the time at the start of the first call
5326 to select(). If more time than the timeout has elapsed, the interrupt is
5327 treated as a timeout.
5329 5. Some internal re-factoring in preparation for the addition of Sieve
5330 extensions (by MH). In particular, the "personal" test is moved to a
5331 separate function, and given an option for scanning Cc: and Bcc: (which is
5332 not set for Exim filters).
5334 6. When Exim created an email address using the login of the caller as the
5335 local part (e.g. when creating a From: or Sender: header line), it was not
5336 quoting the local part when it contained special characters such as @.
5338 7. Installed new OpenBSD configuration files.
5340 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5341 try to make them clearer.
5343 9. Callout options, other than the timeout value, were being ignored when
5344 verifying sender addresses in header lines. For example, when using
5346 verify = header_sender/callout=no_cache
5348 the cache was (incorrectly) being used.
5350 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5351 only the Cygwin environment.
5353 11. When return_path_on_delivery was set as a log selector, if different remote
5354 addresses in the same message used different return paths and parallel
5355 remote delivery occurred, the wrong values would sometimes be logged.
5356 (Whenever a remote delivery process finished, the return path value from
5357 the most recently started remote delivery process was logged.)
5359 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5360 lines when AUTH and/or TLS are in use. This is the "received protocol"
5361 field. Exim used to use "asmtp" for authenticated SMTP, without any
5362 indication (in the protocol name) for TLS use. Now it follows the RFC and
5363 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5364 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5365 names appear in log lines as well as in Received: header lines.
5367 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5368 extensions, and comparison tests, and to fix some bugs.
5370 14. Changes to the "personal" filter test:
5372 (1) The test was buggy in that it was just doing the equivalent of
5373 "contains" tests on header lines. For example, if a user's address was
5374 anne@some.where, the "personal" test would incorrectly be true for
5376 To: susanne@some.where
5378 This test is now done by extracting each address from the header in turn,
5379 and checking the entire address. Other tests that are part of "personal"
5380 are now done using regular expressions (for example, to check local parts
5381 of addresses in From: header lines).
5383 (2) The list of non-personal local parts in From: addresses has been
5384 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5385 taken from the Sieve specification recommendations.
5387 (3) If the message contains any header line starting with "List-" it is
5388 treated as non-personal.
5390 (4) The test for "circular" in the Subject: header line has been removed
5391 because it now seems ill-conceived.
5393 15. Minor typos in src/EDITME comments corrected.
5395 16. Installed latest exipick from John Jetmore.
5397 17. If headers_add on a router specified a text string that was too long for
5398 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5399 of string_sprintf() is now avoided.
5401 18. $message_body_size was not set (it was always zero) when running the DATA
5402 ACL and the local_scan() function.
5404 19. For the "mail" command in an Exim filter, no default was being set for
5405 the once_repeat time, causing a random time value to be used if "once" was
5406 specified. (If the value happened to be <= 0, no repeat happened.) The
5407 default is now 0s, meaning "never repeat". The "vacation" command was OK
5408 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5409 (I found it when inspecting the code).
5411 20. There is now an overall timeout for performing a callout verification. It
5412 defaults to 4 times the callout timeout, which applies to individual SMTP
5413 commands during the callout. The overall timeout applies when there is more
5414 than one host that can be tried. The timeout is checked before trying the
5415 next host. This prevents very long delays if there are a large number of
5416 hosts and all are timing out (e.g. when the network connections are timing
5417 out). The value of the overall timeout can be changed by specifying an
5418 additional sub-option for "callout", called "maxwait". For example:
5420 verify = sender/callout=5s,maxwait=20s
5422 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5423 to the end before writing, but this should make it even safer).
5425 22. Exim was forgetting that it had advertised PIPELINING for the second and
5426 subsequent messages on an SMTP connection. It was also not resetting its
5427 memory on STARTTLS and an internal HELO.
5429 23. When Exim logs an SMTP synchronization error within a session, it now
5430 records whether PIPELINING has been advertised or not.
5432 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5433 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5434 rather than long int.
5436 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5438 26. Added the never_mail option to autoreply.
5444 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5445 crash if the getsockname() call failed; this can happen if a connection is
5446 closed very soon after it is established. The problem was simply in the
5447 order in which certain operations were done, causing Exim to try to write
5448 to the SMTP stream before it had set up the file descriptor. The bug has
5449 been fixed by making things happen in the correct order.
5455 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5456 before the connection was closed, thus losing the rejection response.
5458 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5459 some early Solaris releases, but causes trouble in current releases where
5460 socklen_t is defined.
5462 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5465 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5468 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5469 that releasing the top of it at the end releases what was used for sub-
5470 expansions (unless the block got too big). However, discard this block if
5471 the first thing is a variable or header, so that we can use its block when
5472 it is dynamic (useful for very large $message_headers, for example).
5474 6. Lookups now cache *every* query, not just the most recent. A new, separate
5475 store pool is used for this. It can be recovered when all lookup caches are
5476 flushed. Lookups now release memory at the end of their result strings.
5477 This has involved some general refactoring of the lookup sources.
5479 7. Some code has been added to the store_xxx() functions to reduce the amount
5480 of flapping under certain conditions.
5482 8. log_incoming_interface used to affect only the <= reception log lines. Now
5483 it causes the local interface and port to be added to several more SMTP log
5484 lines, for example "SMTP connection from", and rejection lines.
5486 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5488 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5490 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5491 could be overwritten at the end of the current message (or the start of a
5492 new message if it was set in a HELO ACL). The value is now preserved for
5493 the duration of the SMTP connection.
5495 12. If a transport had a headers_rewrite setting, and a matching header line
5496 contained an unqualified address, that address was qualified, even if it
5497 did not match any rewriting rules. The underlying bug was that the values
5498 of the flags that permit the existence of unqualified sender and recipient
5499 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5500 non-local messages, and by -bnq for local messages) were not being
5501 preserved with the message after it was received.
5503 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5504 "next input=" as part of the text comprising the host identity instead of
5505 the correct text. The code was using the same buffer for two different
5506 strings. However, depending on which order the printing function evaluated
5507 its arguments, the bug did not always show up. Under Linux, for example, my
5508 test suite worked just fine.
5510 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5511 doesn't work with some older versions of Perl. It has been changed to "my",
5512 which in any case is probably the better facility to use.
5514 15. A really picky compiler found some instances of statements for creating
5515 error messages that either had too many or two few arguments for the format
5518 16. The size of the buffer for calls to the DNS resolver has been increased
5519 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5520 for addresses that have a lot of PTR records. This alleviates a problem; it
5521 does not fully solve it.
5523 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5524 buffer now truncates the list and logs the incident, which is the same
5525 action as happens when Exim is looking up a host name and its aliases.
5526 Previously in this situation something unpredictable would happen;
5527 sometimes it was "internal error: store_reset failed".
5529 18. If a server dropped the connection unexpectedly when an Exim client was
5530 using GnuTLS and trying to read a response, the client delivery process
5531 crashed while trying to generate an error log message.
5533 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5534 string, for example:
5536 warn message = H1: something\nH2: something
5538 the text was added as a single header line from Exim's point of view
5539 though it ended up OK in the delivered message. However, searching for the
5540 second and subsequent header lines using $h_h2: did not work. This has been
5541 fixed. Similarly, if a system filter added multiple headers in this way,
5542 the routers could not see them.
5544 20. Expanded the error message when iplsearch is called with an invalid key to
5545 suggest using net-iplsearch in a host list.
5547 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5548 ACLs are no longer actually imposed (and a message to that effect is
5551 22. If a "gecos" field in a passwd entry contained escaped characters, in
5552 particular, if it contained a \" sequence, Exim got it wrong when building
5553 a From: or a Sender: header from that name. A second bug also caused
5554 incorrect handling when an unquoted " was present following a character
5555 that needed quoting.
5557 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5558 was not being matched caselessly.
5560 24. Arranged for all hyphens in the exim.8 source to be escaped with
5563 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5564 properly. Recipient callout cache records were still being keyed to include
5565 the sender, even when use_sender was set false. This led to far more
5566 callouts that were necessary. The sender is no longer included in the key
5567 when use_sender is false.
5569 26. Added "control = submission" modifier to ACLs.
5571 27. Added the ${base62d: operator to decode base 62 numbers.
5573 28. dnsdb lookups can now access SRV records.
5575 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5576 the configuration file.
5578 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5579 (-v) mode. This makes the output for a verbose queue run more intelligible.
5581 31. Added a use_postmaster feature to recipient callouts.
5583 32. Added the $body_zerocount variable, containing the number of binary zero
5584 bytes in the message body.
5586 33. The time of last modification of the "new" subdirectory is now used as the
5587 "mailbox time last read" when there is a quota error for a maildir
5590 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5592 35. Added +ignore_unknown as a special item in host lists.
5594 36. Code for decoding IPv6 addresses in host lists is now included, even if
5595 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5596 address was recognized as an IP address, but was then not correctly decoded
5597 into binary, causing unexpected and incorrect effects when compared with
5604 1. Very minor rewording of debugging text in manualroute to say "list of
5605 hosts" instead of "hostlist".
5607 2. If verify=header_syntax was set, and a header line with an unqualified
5608 address (no domain) and a large number of spaces between the end of the
5609 name and the colon was received, the reception process suffered a buffer
5610 overflow, and (when I tested it) crashed. This was caused by some obsolete
5611 code that should have been removed. The fix is to remove it!
5613 3. When running in the test harness, delay a bit after writing a bounce
5614 message to get a bit more predictability in the log output.
5616 4. Added a call to search_tidyup() just before forking a reception process. In
5617 theory, someone could use a lookup in the expansion of smtp_accept_max_
5618 per_host which, without the tidyup, could leave open a database connection.
5620 5. Added the variables $recipient_data and $sender_data which get set from a
5621 lookup success in an ACL "recipients" or "senders" condition, or a router
5622 "senders" option, similar to $domain_data and $local_part_data.
5624 6. Moved the writing of debug_print from before to after the "senders" test
5627 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5628 problems for message scanning, either using a data ACL, or using
5629 local_scan() because the Received: header was not generated till after they
5630 were called (in order to set the time as the time of reception completion).
5631 I have revised the way this works. The header is now generated after the
5632 body is received, but before the ACL or local_scan() are called. After they
5633 are run, the timestamp in the header is updated.
5639 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5640 before starting a queue runner without re-exec. This happened only when
5641 deliver_drop_privilege was set or when the Exim user was set to root. The
5642 effect of the bug was that timeouts during subsequent deliveries caused
5643 crashes instead of being properly handled. The handler is now left at its
5644 default (and expected) setting.
5646 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5647 message, again when deliver_drop_privilege is set or Exim is run as root.
5648 The bug described in (1) was not present in this case, but the tidying up
5649 of the other signals was missing. I have made the two cases consistent.
5651 3. The ignore_target_hosts setting on a manualroute router was being ignored
5652 for hosts that were looked up using the /MX notation.
5654 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5657 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5658 operated on the sender address. After changing the $sender_address to <>
5659 for the sender address verify, Exim was re-instated it as the original
5660 (before rewriting) address, but remembering that it had rewritten it, so it
5661 wasn't rewriting it again. This bug also had the effect of breaking the
5662 sender address verification caching when the sender address was rewritten.
5664 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5665 This has been changed so that if the ip literal address matches
5666 ignore_target_hosts, the router declines.
5668 7. Added expansion conditions match_domain, match_address, and match_local_
5669 part (NOT match_host).
5671 8. The placeholder for the Received: header didn't have a length field set.
5673 9. Added code to Exim itself and to exim_lock to test for a specific race
5674 condition that could lead to file corruption when using MBX delivery. The
5675 issue is with the lockfile that is created in /tmp. If this file is removed
5676 after a process has opened it but before that process has acquired a lock,
5677 there is the potential for a second process to recreate the file and also
5678 acquire a lock. This could lead to two Exim processes writing to the file
5679 at the same time. The added code performs the same test as UW imapd; it
5680 checks after acquiring the lock that its file descriptor still refers to
5681 the same named file.
5683 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5684 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5687 11. Added the smtp_active_hostname option. If used, this will typically be made
5688 to depend on the incoming interface address. Because $interface_address is
5689 not set up until the daemon has forked a reception process, error responses
5690 that can happen earlier (such as "too many connections") no longer contain
5693 12. If an expansion in a condition on a "warn" statement fails because a lookup
5694 defers, the "warn" statement is abandoned, and the next ACL statement is
5695 processed. Previously this caused the whole ACL to be aborted.
5697 13. Added the iplsearch lookup type.
5699 14. Added ident_timeout as a log selector.
5701 15. Added tls_certificate_verified as a log selector.
5703 16. Added a global option tls_require_ciphers (compare the smtp transport
5704 option of the same name). This controls incoming TLS connections.
5706 17. I finally figured out how to make tls_require_ciphers do a similar thing
5707 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5708 before starting the TLS session.
5710 18. Tabs are now shown as \t in -bP output.
5712 19. If the log selector return_path_on_delivery was set, Exim crashed when
5713 bouncing a message because it had too many Received: header lines.
5715 20. If two routers both had headers_remove settings, and the first one included
5716 a superfluous trailing colon, the final name in the first list and the
5717 first name in the second list were incorrectly joined into one item (with a
5718 colon in the middle).
5724 1. Added -C and -D options to the exinext utility, mainly to make it easier
5725 to include in the automated testing, but these could be helpful when
5726 multiple configurations are in use.
5728 2. The exinext utility was not formatting the output nicely when there was
5729 an alternate port involved in the retry record key, nor when there was a
5730 message id as well (for retries that were specific to a specific message
5731 and a specific host). It was also confused by IPv6 addresses, because of
5732 the additional colons they contain. I have fixed the IPv4 problem, and
5733 patched it up to do a reasonable job for IPv6.
5735 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5736 delivery, the log line now contains "pipelined" if PIPELINING was used.
5738 4. An SMTP transport process used to panic and die if the bind() call to set
5739 an explicit outgoing interface failed. This has been changed; it is now
5740 treated in the same way as a connect() failure.
5742 5. A reference to $sender_host_name in the part of a conditional expansion
5743 that was being skipped was still causing a DNS lookup. This no longer
5746 6. The def: expansion condition was not recognizing references to header lines
5747 that used bh_ and bheader_.
5749 7. Added the _cache feature to named lists.
5751 8. The code for checking quota_filecount in the appendfile transport was
5752 allowing one more file than it should have been.
5754 9. For compatibility with Sendmail, the command line option
5762 and sets the incoming protocol and host name (for trusted callers). The
5763 host name and its colon can be omitted when only the protocol is to be set.
5764 Note the Exim already has two private options, -pd and -ps, that refer to
5765 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5766 "s", but I don't think that's a major issue.
5768 10. A number of refactoring changes to the code, none of which should affect
5771 (a) The number of logging options was getting close to filling up the
5772 32-bit word that was used as a bit map. I have split them into two classes:
5773 those that are passed in the argument to log_write(), and those that are
5774 only ever tested independently outside of that function. These are now in
5775 separate 32-bit words, so there is plenty of room for expansion again.
5776 There is no change in the user interface or the logging behaviour.
5778 (b) When building, for example, log lines, the code previously used a
5779 macro that called string_cat() twice, in order to add two strings. This is
5780 not really sufficiently general. Furthermore, there was one instance where
5781 it was actually wrong because one of the argument was used twice, and in
5782 one call a function was used. (As it happened, calling the function twice
5783 did not affect the overall behaviour.) The macro has been replaced by a
5784 function that can join an arbitrary number of extra strings onto a growing
5787 (c) The code for expansion conditions now uses a table and a binary chop
5788 instead of a serial search (which was left over from when there were very
5789 few conditions). Also, it now recognizes conditions like "pam" even when
5790 the relevant support is not compiled in: a suitably worded error message is
5791 given if an attempt is made to use such a condition.
5793 11. Added ${time_interval:xxxxx}.
5795 12. A bug was causing one of the ddress fields not to be passed back correctly
5796 from remote delivery subprocesses. The field in question was not being
5797 subsequently used, so this caused to problems in practice.
5799 13. Added new log selectors queue_time and deliver_time.
5801 14. Might have fixed a bug in maildirsizefile handling that threw up
5802 "unexpected character" debug warnings, and recalculated the data
5803 unnecessarily. In any case, I expanded the warning message to give more
5806 15. Added the message "Restricted characters in address" to the statements in
5807 the default ACL that block characters like @ and % in local parts.
5809 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5810 Three changes have been made:
5812 (a) There was a serious bug; a negative response to MAIL caused the whole
5813 recipient domain to be cached as invalid, thereby blocking all messages
5814 to all local parts at the same domain, from all senders. This bug has
5815 been fixed. The domain is no longer cached after a negative response to
5816 MAIL if the sender used is not empty.
5818 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5821 (c) A new callout option, "use_sender" has been added for people who want
5822 the modified behaviour.
5828 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5831 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5832 indeed breaks things for older releases.
5834 3. Added additional logging to the case where there is a problem reading data
5835 from a filter that is running in a subprocess using a pipe, in order to
5836 try to track down a specific problem.
5838 4. Testing facility fudge: when running in the test harness and attempting
5839 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5840 getting "No route to host". Convert this to a timeout.
5842 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5845 6. Some OS don't have socklen_t but use size_t instead. This affects the
5846 fifth argument of getsockopt() amongst other things. This is now
5847 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5848 can be set for individual OS. I have set it for SunOS5, OSF1, and
5849 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5850 some earlier ones do not.
5852 7. Change 4.30/15 was not doing the test caselessly.
5854 8. The standard form for an IPv6 address literal was being rejected by address
5855 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5856 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5857 this, as well as the form without the "IPv6" on the front (but only when
5858 address literals are enabled, of course).
5860 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5862 10. Exim crashed if a message with an empty sender address specified by -f
5863 encountered a router with an errors_to setting. This could be provoked only
5864 by a command such as
5868 where an empty string was supplied; "<>" did not hit this bug.
5870 11. Installed PCRE release 4.5.
5872 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5873 remained set. It is now erased.
5875 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5876 times from message ids (which are base 36 rather than the normal 62).
5878 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5879 were being counted as actual protocol errors, and logged if the log
5880 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5881 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5882 and DATA following a set of rejected RCPTs do not count as protocol errors.
5883 In other words, Exim assumes they were pipelined, though this may not
5884 actually be the case. Of course, in all cases the client gets an
5885 appropriate error code.
5887 15. If a lookup fails in an ACL condition, a message about the failure may
5888 be available; it is used if testing the ACL cannot continue, because most
5889 such messages specify what the cause of the deferral is. However, some
5890 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5891 that caused an old message to be retained and used if a later statement
5892 caused a defer, replacing the real cause of the deferral.
5894 16. If an IP address had so many PTR records that the DNS lookup buffer
5895 was not large enough to hold them, Exim could crash while trying to process
5896 the truncated data. It now detects and logs this case.
5898 17. Further to 4.21/58, another change has been made: if (and only if) the
5899 first line of a message (the first header line) ends with CRLF, a bare LF
5900 in a subsequent header line has a space inserted after it, so as not to
5901 terminate the header.
5903 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5904 unnecessarily, used atoi() instead of strtol(), and didn't check the
5905 termination when getting file sizes from file names by regex.
5907 19. Completely re-implemented the support for maildirsize files, in the light
5908 of a number of problems with the previous contributed implementation
5909 (4.30/29). In particular:
5911 . If the quota is zero, the maildirsize file is maintained, but no quota is
5914 . If the maildir directory does not exist, it is created before any attempt
5915 to write a maildirsize file.
5917 . The quota value in the file is just a cache; if the quota is changed in
5918 the transport, the new value overrides.
5920 . A regular expression is available for excluding directories from the
5923 20. The autoreply transport checks the characters in options that define the
5924 message's headers; it allows continued headers, but it was checking with
5925 isspace() after an embedded newline instead of explicitly looking for a
5928 21. If all the "regular" hosts to which an address was routed had passed their
5929 expiry times, and had not reached their retry times, the address was
5930 bounced, even if fallback hosts were defined. Now Exim should go on to try
5933 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5934 equivalent code in the SMTP transport. Some hosts send humungous responses
5935 to HELO/EHLO, more than 1024 it seems.
5937 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5938 gives compiler warnings in some environments. I've now done it "properly",
5941 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5942 (because of problems with the built-in one) was declared to return uschar *
5943 instead of char *, causing compiler failure.
5945 25. Fixed a file descriptor leak when processing alias/forward files.
5947 26. Fixed a minor format string issue in dbfn.c.
5949 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5951 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5952 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5955 29. When a sender address is verified, it is cached, to save repeating the test
5956 when there is more than one recipient in a message. However, when the
5957 verification involves a callout, it is possible for different callout
5958 options to be set for different recipients. It is too complicated to keep
5959 track of this in the cache, so now Exim always runs a verification when a
5960 callout is required, relying on the callout cache for the optimization.
5961 The overhead is duplication of the address routing, but this should not be
5964 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5965 to be verified with callout=postmaster, and the main callout worked but the
5966 postmaster check failed, the verification correctly failed. However, if a
5967 subsequent RCPT command asked for sender verification *without* the
5968 postmaster check, incorrect caching caused this verification also to fail,
5971 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5972 it was not caching the DNS options (qualify_single, search_parents) that
5973 were used when the lookup failed. A subsequent lookup with different
5974 options therefore always gave the same answer, though there were cases
5975 where it should not have. (Example: a "domains = !$mx_any" option on a
5976 dnslookup router: the "domains" option is always processed without any
5977 widening, but the router might have qualify_single set.) Now Exim uses the
5978 cached value only when the same options are set.
5980 32. Added John Jetmore's "exipick" utility to the distribution.
5982 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5983 than a timeout (e.g. a certificate is required, and is not provided), an
5984 Exim server now closes the connection immediately. Previously it waited for
5985 the client to close - but if the client is SSL, it seems that they each
5986 wait for each other, leading to a delay before one of them times out.
5988 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5989 maintained 0.8.x compatibility because I don't think many are using it, and
5990 it is clearly obsolete.
5992 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5995 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5996 client certificate was expired. A simple patch fixes this, though I don't
5997 understand the full logic of why the verify callback is called multiple
6000 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6001 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6002 which causes problems with some clients (such as the Certicom SSL Plus
6003 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6004 disables the coutermeasure allowing Eudora to connect."
6006 38. Exim was not checking that a write() to a log file succeeded. This could
6007 lead to Bad Things if a log got too big, in particular if it hit a file
6008 size limit. Exim now panics and dies if it cannot write to a log file, just
6009 as it does if it cannot open a log file.
6011 39. Modified OS/Makefile-Linux so that it now contains
6013 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6015 The two -D definitions ensure that Exim is compiled with large file
6016 support, which makes it possible to handle log files that are bigger than
6019 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6020 instance) a domain was checked against a named list that involved a lookup,
6021 causing $domain_data to be set, then another domain was checked against the
6022 same list, then the first domain was re-checked, the value of $domain_data
6023 after the final check could be wrong. In particular, if the second check
6024 failed, it could be set empty. This bug probably also applied to
6027 41. The strip_trailing_dot option was not being applied to the address given
6028 with the -f command-line option.
6030 42. The code for reading a message's header from the spool was incrementing
6031 $received_count, but never initializing it. This meant that the value was
6032 incorrect (doubled) while delivering a message in the same process in which
6033 it was received. In the most common configuration of Exim, this never
6034 happens - a fresh exec is done - but it can happen when
6035 deliver_drop_privilege is set.
6037 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6038 it now includes up to 150 characters of the unexpected data in the log
6041 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6042 and building data strings. The size of both of these buffers was 10 000
6043 bytes - far larger than anybody would *ever* want, thought I. Needless to
6044 say, somebody hit the limit. I have increased the maximum line length to
6045 20 000 and the maximum data length of concatenated lines to 100 000. I have
6046 also fixed two bugs, because there was no checking on these buffers. Tsk,
6047 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6048 buffer is too small.
6050 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6051 lsearch lookups. Now it does.
6053 46. When parsing a route_list item in a manualroute router, a fixed-length
6054 buffer was used for the list of hosts. I made this 1024 bytes long,
6055 thinking that nobody would ever have a list of hosts that long. Wrong.
6056 Somebody had a whole pile of complicated expansion conditions, and the
6057 string was silently truncated, leading to an expansion error. It turns out
6058 that it is easier to change to an unlimited length (owing to other changes
6059 that have happened since this code was originally written) than to build
6060 structure for giving a limitation error. The length of the item that
6061 expands into the list of hosts is now unlimited.
6063 47. The lsearch lookup could not handle data where the length of text line was
6064 more than 4095 characters. Such lines were truncated, leading to shortened
6065 data being returned. It should now handle lines of any length.
6067 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6068 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6071 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6072 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6073 no longer happens in comment lines. A list of replacements is now placed
6074 at the head of all of the source files, except those whose only change is
6075 to replace PERL_COMMAND in the very first #! line.
6077 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6078 messages on the queue, with a bottom-up merge sort, using code contributed
6079 by Michael Haardt. This should make operations like -bp somewhat faster on
6080 large queues. It won't affect queue runners, except when queue_run_in_order
6083 51. Installed eximstats 1.31 in the distribution.
6085 52. Added support for SRV lookups to the dnslookup router.
6087 53. If an ACL referred to $message_body or $message_body_end, the value was not
6088 reset for any messages that followed in the same SMTP session.
6090 54. The store-handling optimization for building very long strings was not
6091 differentiating between the different store pools. I don't think this
6092 actually made any difference in practice, but I've tidied it.
6094 55. While running the routers to verify a sender address, $sender_address
6095 was still set to the sender address. This is wrong, because when routing to
6096 send a bounce to the sender, it would be empty. Therefore, I have changed
6097 it so that, while verifying a sender address, $sender_address is set to <>.
6098 (There is no change to what happens when verifying a recipient address.)
6100 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6101 target A or AAAA records (if not already returned) without resetting the
6102 qualify_single or search_parents options of the DNS resolver. These are
6103 inappropriate in this case because the targets of MX and SRV records must
6104 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6105 target that, when qualified, matched something in the local domain. These
6106 two options are now turned off when doing these lookups.
6108 57. It seems that at least some releases of Reiserfs (which does not have the
6109 concept of a fixed number of inodes) returns zero and not -1 for the
6110 number of available inodes. This interacted badly with check_spool_inodes,
6111 which assumed that -1 was the "no such thing" setting. What I have done is
6112 to check that the total number of inodes is greater than zero before doing
6113 the test of how many are available.
6115 58. When a "warn" ACL statement has a log_message modifier, the message is
6116 remembered, and not repeated. This is to avoid a lot of repetition when a
6117 message has many recipients that cause the same warning to be written.
6118 However, Exim was preserving the list of already written lines for an
6119 entire SMTP session, which doesn't seem right. The memory is now reset if a
6120 new message is started.
6122 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6123 debugging output unless log_rewrite was also set.
6125 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6126 of (int)(handle) when we know that handle contains (void *)(-1).
6128 61. The Exim daemon panic-logs an error return when it closes the incoming
6129 connection. However "connection reset by peer" seems to be common, and
6130 isn't really an error worthy of noting specially, so that particular error
6133 62. When Exim is trying to find all the local interfaces, it used to panic and
6134 die if the ioctl to get the interface flags failed. However, it seems that
6135 on at least one OS (Solaris 9) it is possible to have an interface that is
6136 included in the list of interfaces, but for which you get a failure error
6137 for this call. This happens when the interface is not "plumbed" into a
6138 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6139 failure of the "get flags" call assumes that the interface is down.
6141 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6142 makes life easier for people who are doing arithmetic on fields extracted
6143 from dates, where you often get leading zeros that should not be
6144 interpreted as octal.
6146 64. Added qualify_domain to the redirect router, to override the global
6149 65. If a pathologically long header line contained very many addresses (the
6150 report of this problem mentioned 10 000) and each of them was rewritten,
6151 Exim could use up a very large amount of memory. (It kept on making new
6152 copies of the header line as it rewrote, and never released the old ones.)
6153 At the expense of a bit more processing, the header rewriting function has
6154 been changed so that it no longer eats memory in this way.
6156 66. The generation of the Received: header has been moved from the time that a
6157 message starts to be received, to the time that it finishes. The timestamp
6158 in the Received: header should now be very close to that of the <= log
6159 line. There are two side-effects of this change:
6161 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6162 logged header lines no longer include the local Received: line, because
6163 it has not yet been created. The same applies to a copy of the message
6164 that is returned to a non-SMTP sender when a message is rejected.
6166 (b) When a filter file is tested using -bf, no additional Received: header
6167 is added to the test message. After some thought, I decided that this
6170 This change does not affect the value of $received_for. It is still set
6171 after address rewriting, but before local_scan() is called.
6173 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6175 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6176 gave an unhelpful panic error message, and a defer error. I have managed to
6177 change this behaviour so that it now rejects any supplied certificate,
6178 which seems right, as the list of acceptable certificates is empty.
6180 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6181 gave an unhelpful defer error. I have not managed to make this reject any
6182 supplied certificates, but the error message it gives is "no certificate
6183 supplied", which is not helpful.
6185 70. exigrep's output now also includes lines that are not associated with any
6186 message, but which match the given pattern. Implemented by a patch from
6187 Martin Sluka, which also tidied up the Perl a bit.
6189 71. Recipient callout verification, like sender verification, was using <> in
6190 the MAIL FROM command. This isn't really the right thing, since the actual
6191 sender may affect whether the remote host accepts the recipient or not. I
6192 have changed it to use the actual sender in the callout; this means that
6193 the cache record is now keyed on a recipient/sender pair, not just the
6194 recipient address. There doesn't seem to be a real danger of callout loops,
6195 since a callout by the remote host to check the sender would use <>.
6196 [SEE ABOVE: changed after hitting problems.]
6198 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6199 temporary errors. However, in the case of such a code being given after
6200 the end of a data transmission (i.e. after ".") Exim was failing to write
6201 a retry record for the message. (Yes, there was some broken host that was
6202 actually sending 8xx at this point.)
6204 73. An unknown lookup type in a host list could cause Exim to panic-die when
6205 the list was checked. (An example that provoked this was putting <; in the
6206 middle of a list instead of at the start.) If this happened during a DATA
6207 ACL check, a -D file could be left lying around. This kind of configuration
6208 error no longer causes Exim to die; instead it causes a defer error. The
6209 incident is still logged to the main and panic logs.
6211 74. Buglet left over from Exim 3 conversion. The message "too many messages
6212 in one connection" was written to the rejectlog but not the mainlog, except
6213 when address rewriting (yes!) was being logged.
6215 75. Added write_rejectlog option.
6217 76. When a system filter was run not as root (that is, when system_filter_user
6218 was set), the values of the $n variables were not being returned to the
6219 main process; thus, they were not subsequently available in the $sn
6222 77. Added +return_path_on_delivery log selector.
6224 78. A connection timeout was being treated differently from recipients deferred
6225 when testing hosts_max_try with a message that was older than the host's
6226 retry timeout. (The host should not be counted, thus allowing all hosts to
6227 be tried at least once before bouncing.) This may have been the cause of an
6228 occasionally reported bug whereby a message would remain on the queue
6229 longer than the retry timeout, but would be bounced if a delivery was
6230 forced. I say "may" because I never totally pinned down the problem;
6231 setting up timeout/retry tests is difficult. See also the next item.
6233 79. The ultimate address timeout was not being applied to errors that involved
6234 a combination of host plus message (for example, a timeout on a MAIL
6235 command). When an address resolved to a number of possible hosts, and they
6236 were not all tried for each delivery (e.g. because of hosts_max_try), a
6237 message could remain on the queue longer than the retry timeout.
6239 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6242 81. Fixed an obscure SMTP outgoing bug which required at least the following
6243 conditions: (a) there was another message waiting for the same server;
6244 (b) the server returned 5xx to all RCPT commands in the first message so
6245 that the message was not completed; (c) the server dropped the connection
6246 or gave a negative response to the RSET that Exim sends to abort the
6247 transaction. The observed case was a dropped connection after DATA that had
6248 been sent in pipelining mode. That is, the server had advertised PIPELINING
6249 but was not implementing it correctly. The effect of the bug was incorrect
6250 behaviour, such as trying another host, and this could lead to a crash.
6256 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6257 and daemon.c were passed as pointers to ints; they should have been
6258 pointers to socklen_t variables (which are typically unsigned ints).
6260 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6263 3. Fixed a really odd bug that affected only the testing scheme; patching a
6264 certain fixed string in the binary changed the value of another string that
6265 happened to be identical to the end of the original first string.
6267 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6268 name", it returns that address as the IP address. On some operating
6269 systems (e.g. Solaris), it also passes back the IP address string as the
6270 "host name". However, on others (e.g. Linux), it passes back an empty
6271 string. Exim wasn't checking for this, and was changing the host name to an
6272 empty string, assuming it had been canonicalized.
6274 5. Although rare, it is permitted to have more than one PTR record for a given
6275 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6276 all the names associated with an address, because they do in Solaris.
6277 However, it seems that they do not in Linux for data that comes from the
6278 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6279 I found this out when I moved to a new Linux workstation and tried to run
6280 the Exim test suite.
6282 To get round this problem I have changed the code so that it now does its
6283 own call to the DNS to look up PTR records when searching for a host name.
6284 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6285 addresses that are only in /etc/hosts are still found.
6287 This behaviour is, however, controlled by an option called host_lookup_
6288 order, which defaults to "bydns:byaddr". If people want to use the other
6289 order, or indeed, just use one or the other means of lookup, they can
6290 specify it in this variable.
6292 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6293 some operating systems, this comes back from gethostbyaddr() as an empty
6294 string, and this is what Exim used to test for. However, it seems that in
6295 other systems, "." is yielded. Exim now tests for this case too.
6297 7. The values of check_spool_space and check_log_space are now held internally
6298 as a number of kilobytes instead of an absolute number of bytes. If a
6299 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6300 kilobyte. This means that much larger values can be stored.
6302 8. Exim monitor: an attempt to get the action menu when not actually pointing
6303 at a message produces an empty menu entitled "No message selected". This
6304 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6305 no entries in it ("Shell widget menu has zero width and/or height"). So I
6306 have added a single, blank menu entry in this case.
6308 9. Added ${quote_local_part.
6310 10. MIME decoding is now applied to the contents of Subject: header lines when
6313 11. Now that a reference to $sender_host_address automatically causes a reverse
6314 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6315 host lookup before query-style lookups in lists that might use this
6316 variable. This has therefore been abolished, and the "net-" prefix is no
6317 longer necessary for query-style lookups.
6319 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6320 have been a typo for LFLAGS, so it has been changed.
6322 13. The install script calls Exim with "-C /dev/null" in order to find the
6323 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6324 to be output. However, since Exim outputs its version number before the
6325 error, it didn't break the script. It just looked ugly. I fixed this by
6326 always allowing "-C /dev/null" if the caller is root.
6328 14. Ignore overlarge ACL variable number when reading spool file - insurance
6329 against a later release with more variables having written the file.
6331 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6332 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6333 this, as well as the form without the "IPv6" on the front.
6335 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6336 OS/Makefile-Darwin file.
6338 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6339 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6341 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6342 variable. However, some OS use ints for inodes. I've added cast to long int
6343 to get rid of the compiler warning.
6345 19. I had forgotten to lock out "/../" in configuration file names when
6346 ALT_CONFIG_PREFIX was set.
6348 20. Routers used for verification do not need to specify transports. However,
6349 if such a router generated a host list, and callout was configured, Exim
6350 crashed, because it could not find a port number from the (non-existent)
6351 transport. It now assumes port 25 in this circumstance.
6353 21. Added the -t option to exigrep.
6355 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6356 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6357 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6358 them is set, LOOKUP_LSEARCH is forced.
6360 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6362 24. Added sender and host information to the "rejected by local_scan()" log
6363 line; previously there was no indication of these.
6365 25. Added .include_if_exists.
6367 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6368 Linux. It turns out that not all file systems support this. Apparently some
6369 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6370 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6371 not supported on the file descriptor, is now ignored when Exim is trying to
6372 sync a directory. This applies only to Linux.
6374 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6376 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6377 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6380 29. Added support for maildirsize files from supplied patch (modified a bit).
6382 30. The use of :fail: followed by an empty string could lead Exim to respond to
6383 sender verification failures with (e.g.):
6385 550 Verification failed for <xxx>
6386 550 Sender verify failed
6388 where the first response line was missing the '-' that indicates it is not
6389 the final line of the response.
6391 31. The loop for finding the name of the user that called Exim had a hardwired
6392 limit of 10; it now uses the value of finduser_retries, which is used for
6393 all other user lookups.
6395 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6398 33. Exim was neglecting to zero errno before one call of strtol() when
6399 expanding a string and expecting an integer value. On some systems this
6400 resulted in spurious "integer overflow" errors. Also, it was casting the
6401 result into an int without checking.
6403 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6404 did not work. The code looks as if it has *never* worked, though it appears
6405 to have been documented since at least release 1.62. I have made it work.
6407 35. The "timeout_DNS" error in retry rules, also documented since at least
6408 1.62, also never worked. As it isn't clear exactly what this means, and
6409 clearly it isn't a major issue, I have abolished the feature by treating it
6410 as "timeout", and writing a warning to the main and panic logs.
6412 36. The display of retry rules for -brt wasn't always showing the error code
6415 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6416 timeout_connect_A, timeout_connect_MX.
6418 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6419 to the empty sender.
6421 39. The daemon was not analysing the content of -oX till after it had closed
6422 stderr and disconnected from the controlling terminal. This meant that any
6423 syntax errors were only noted on the panic log, and the return code from
6424 the command was 0. By re-arranging the code a little, I've made the
6425 decoding happen first, so such errors now appear on stderr, and the return
6426 code is 1. However, the actual setting up of the sockets still happens in
6427 the disconnected process, so errors there are still only recorded on the
6430 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6431 connections (as happens on some IP stacks) was logged at start up time as
6432 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6433 it from "IPv6 and IPv4", which means that two separate sockets are being
6436 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6437 says whether AF_INET or AF_INET6 was passed as an argument.
6439 42. Exiwhat output was messed up when time zones were included in log
6442 43. Exiwhat now gives more information about the daemon's listening ports,
6443 and whether -tls-on-connect was used.
6445 44. The "port" option of the smtp transport is now expanded.
6447 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6448 silently ignored. Now an error message is written to the main and panic
6451 46. There's a new ACL modifier called "logwrite" which writes to a log file
6452 as soon as it is encountered.
6454 47. Added $local_user_uid and $local_user_gid at routing time.
6456 48. Exim crashed when trying to verify a sender address that was being
6459 49. Exim was recognizing only a space character after ".include". It now also
6460 recognizes a tab character.
6462 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6463 extracting the relevant information from the specification. The man page no
6464 longer contains scrambled data for the -d option, and I've added a section
6465 at the front about calling Exim under different names.
6467 51. Added "extra_headers" argument to the "mail" command in filter files.
6469 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6472 53. Installed eximstats 1.29.
6474 54. Added transport_filter_timeout as a generic transport option.
6476 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6477 Cc: header lines. This was required by RFC 822, but it not required by RFC
6480 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6481 incoming messages that did not have them. Now it does so only if the
6482 message originates locally, that is, if there is no associated remote host
6483 address. When Resent- header lines are present, this applies to the Resent-
6484 lines rather than the non-Resent- lines.
6486 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6487 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6489 58. Messages for configuration errors now include the name of the main
6490 configuration file - useful now that there may be more than one file in a
6491 list (.included file names were always shown).
6493 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6494 for those rare installations that do not start the daemon as root or run it
6495 setuid root. I've cut out the call to initgroups() if the daemon is not
6498 60. The Exim user and group can now be bound into the binary as text strings
6499 that are looked up at the start of Exim's processing.
6501 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6503 62. Added $mailstore_basename variable.
6505 63. Installed patch to sieve.c from Michael Haardt.
6507 64. When Exim failed to open the panic log after failing to open the main log,
6508 the original message it was trying to log was written to stderr and debug
6509 output, but if they were not available (the usual case in production), it
6510 was lost. Now it is written to syslog before the two lines that record the
6511 failures to open the logs.
6513 65. Users' Exim filters run in subprocesses under the user's uid. It is
6514 possible for a "deliver" command or an alias in a "personal" command to
6515 provoke an address rewrite. If logging of address rewriting is configured,
6516 this fails because the process is not running as root or exim. There may be
6517 a better way of dealing with this, but for the moment (because 4.30 needs
6518 to be released), I have disabled address rewrite logging when running a
6519 filter in a non-root, non-exim process.
6525 1. The buildconfig auxiliary program wasn't quoting the value set for
6526 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6527 not defined. This bug was masked in 4.22 by the effect that was fixed in
6530 2. Some messages that were rejected after a message id was allocated were
6531 shown as "incomplete" by exigrep. It no longer does this for messages that
6532 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6534 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6535 have allow_domain_literals set, the ID did not get logged in the <= line.
6536 Domain literals are now always recognized in Message-ID: header lines.
6538 4. The first argument for a ${extract expansion item is the key name or field
6539 number. Leading and trailing spaces in this item were not being ignored,
6540 causing some misleading effects.
6542 5. When deliver_drop_privilege was set, single queue runner processes started
6543 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6544 same command in the process it spins off) were not dropping privilege.
6546 6. When the daemon running as "exim" started a queue runner, it always
6547 re-executed Exim in the spun-off process. This is a waste of effort when
6548 deliver_drop_privilege is set. The new process now just calls the
6549 queue-runner function directly.
6555 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6558 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6559 ignored. Though the use of -oP was forcing the writing of a pid file, it
6560 was always written to the default place.
6562 3. If the message "no IP address found for host xxxx" is generated during
6563 incoming verification, it is now followed by identification of the incoming
6564 connection (so you can more easily find what provoked it).
6566 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6568 5. Added some features to "harden" Exim a bit more against certain attacks:
6570 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6571 be put in Local/Makefile. This is like the never_users runtime option,
6572 but it cannot be overridden. The default setting is "root".
6574 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6575 prefix string with which any file named in a -C command line option
6578 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6579 is retained for -C and -D only if the caller of Exim is root. Without
6580 it, the exim user may also use -C and -D and retain privilege.
6582 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6583 command line option is disabled.
6585 6. Macro names set by the -D option must start with an upper case letter, just
6586 like macro names defined in the configuration file.
6588 7. Added "dereference=" facility to LDAP.
6590 8. Two instances of the typo "uknown" in the source files are fixed.
6592 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6593 the Configure-Makefile script screwed up while processing it.
6595 10. Incorporated PCRE 4.4.
6597 11. The SMTP synchronization check was not operating right at the start of an
6598 SMTP session. For example, it could not catch a HELO sent before the client
6599 waited for the greeting. There is now a check for outstanding input at the
6600 point when the greeting is written. Because of the duplex, asynchronous
6601 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6602 way, but not yet received, when the check is performed.
6604 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6605 on TCP/IP sockets, because this apparently causes some broken clients to
6608 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6609 unchanged) from the Cygwin maintainer.
6611 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6612 was supported instead of testing for mbx. Effectively a typo.
6614 15. The spa authenticator server code was not checking that the input it
6615 received was valid base64.
6617 16. The debug output line for the "set" modifier in ACLs was not showing the
6618 name of the variable that was being set.
6620 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6622 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6623 lookup on its own. Something else was needed to trigger the lookup. For
6624 example, a match in host_lookup or the need for a host name in a host list.
6625 Now, if $sender_host_name is referenced and the host name has not yet been
6626 looked up, a lookup is performed. If the lookup fails, the variable remains
6627 empty, and $host_lookup_failed is set to "1".
6629 19. Added "eqi" as a case-independent comparison operator.
6631 20. The saslauthd authentication condition could segfault if neither service
6632 nor realm was specified.
6634 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6635 error message that was logged was misleading, and incoming SMTP
6636 connections were dropped. The message is now more accurate, and temporary
6637 errors are given to SMTP connections.
6639 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6640 (except RSET) with a 421 error, until QUIT is received. However, it was
6641 failing to send a response to QUIT.
6643 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6644 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6645 ACL is now run after the helo_try_verify_hosts code.
6647 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6648 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6649 case-independent, so other case variants are also recognized). Apparently
6650 some systems use these upper case variants.
6652 25. If more than two messages were waiting for the same host, and a transport
6653 filter was specified for the transport, Exim sent two messages over the
6654 same TCP/IP connection, and then failed with "socket operation on non-
6655 socket" when it tried to send the third.
6657 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6659 27. The extern definition of crypt16() in expand.c was not being excluded when
6660 the OS had its own crypt16() function.
6662 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6663 as a preferred synonym for return_size_limit, both as an option and as an
6666 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6668 30. Changed the default configuration ACL to relax the local part checking rule
6669 for addresses that are not in any local domains. For these addresses,
6670 slashes and pipe symbols are allowed within local parts, but the sequence
6671 /../ is explicitly forbidden.
6673 31. SPA server authentication was not clearing the challenge buffer before
6676 32. log_message in a "warn" ACL statement was writing to the reject log as
6677 well as to the main log, which contradicts the documentation and doesn't
6678 seem right (because no rejection is happening). So I have stopped it.
6680 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6681 However, I am unable to do any testing of this.
6683 34. Fixed an infelicity in the appendfile transport. When checking directories
6684 for a mailbox, to see if any needed to be created, it was accidentally
6685 using path names with one or more superfluous leading slashes; tracing
6686 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6688 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6689 contents are added to the log line that is written for every discarded
6690 recipient. (Previously a log_message setting was ignored.)
6692 36. The ${quote: operator now quotes the string if it is empty.
6694 37. The install script runs exim in order to find its version number. If for
6695 some reason other than non-existence or emptiness, which it checks, it
6696 could not run './exim', it was installing it with an empty version number,
6697 i.e. as "exim-". This error state is now caught, and the installation is
6700 38. An argument was missing from the function that creates an error message
6701 when Exim fails to connect to the socket for saslauthd authentication.
6702 This could cause Exim to crash, or give a corrupted message.
6704 39. Added isip, isip4, and isip6 to ${if conditions.
6706 40. The ACL variables $acl_xx are now saved with the message, and can be
6707 accessed later in routers, transports, and filters.
6709 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6710 strings in the file are not string-expanded.
6712 42. If a MAIL command specified a SIZE value that was too large to fit into an
6713 int variable, the check against message_size_limit failed. Such values are
6714 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6715 day this will have to be increased, but I don't think I want to be around
6716 when emails are that large.
6723 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6724 iconv() is not standard in FreeBSD.
6726 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6727 IPv6 enabled. The observed symptom was a segmentation fault on return from
6728 the function os_common_find_running_interfaces() in src/os.c.
6730 3. In the check_special_case() function in daemon.c I had used "errno" as an
6731 argument name, which causes warnings on some systems. This was basically a
6732 typo, since it was named "eno" in the comments!
6734 4. The code that waits for the clock to tick (at a resolution of some fraction
6735 of a second) so as to ensure message-id uniqueness was always waiting for
6736 at least one whole tick, when it could have waited for less. [This is
6737 almost certainly not relevant at current processor speeds, where it is
6738 unlikely to ever wait at all. But we try to future-proof.]
6740 5. The function that sleeps for a time interval that includes fractions of a
6741 second contained a race. It did not block SIGALRM between setting the
6742 timer, and suspending (a couple of lines later). If the interval was short
6743 and the sigsuspend() was delayed until after it had expired, the suspension
6744 never ended. On busy systems this could lead to processes getting stuck for
6747 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6748 process, before it forks any delivery processes. The open lookup caching
6749 mechanism meant that the open file or database connection was passed into
6750 the delivery process. The problem was that delivery processes always tidy
6751 up cached lookup data. This could cause a problem for the next delivery
6752 process started by the queue runner, because the external queue runner
6753 process does not know about the closure. So the next delivery process
6754 still has data in the lookup cache. In the case of a file lookup, there was
6755 no problem because closing a file descriptor in a subprocess doesn't affect
6756 the parent. However, if the lookup was caching a connection to a database,
6757 the connection was closed, and the second delivery process was likely to
6758 see errors such as "PGSQL: query failed: server closed the connection
6759 unexpectedly". The problem has been fixed by closing all cached lookups
6760 in a queue runner before running a delivery process.
6762 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6763 seem to have the "const" qualifier which it has on other OS. I've
6766 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6767 *of the same type* (client or server) with the same public name that an
6768 error should be diagnosed.
6770 9. When Exim looked up a host name for an IP address, but failed to find the
6771 original IP address when looking up the host name (a safety check), it
6772 output the message "<ip address> does not match any IP for NULL", which was
6773 confusing, to say the least. The bug was that the host name should have
6774 appeared instead of "NULL".
6776 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6777 user that is built into the binary, and the -C or -D options is used, root
6778 privilege is dropped before the configuration file is read. In addition,
6779 logging is switched to stderr instead of the normal log files. If the
6780 configuration then re-defines the Exim user, the unprivileged environment
6781 is probably not what is expected, so Exim logs a panic warning message (but
6784 However, if deliver_drop_privilege is set, the unprivileged state may well
6785 be exactly what is intended, so the warning has been cut out in that case,
6786 and Exim is allowed to try to write to its normal log files.
6792 1. smtp_return_error_details was not giving details for temporary sender
6793 or receiver verification errors.
6795 2. Diagnose a configuration error if two authenticators have the same public
6798 3. Exim used not to create the message log file for a message until the first
6799 delivery attempt. This could be confusing when incoming messages were held
6800 for policy or load reasons. The message log file is now created at the time
6801 the message is received, and an initial "Received" line is written to it.
6803 4. The automatically generated man page for command line options had a minor
6804 bug that caused no ill effects; however, a more serious problem was that
6805 the procedure for building the man page automatically didn't always
6806 operate. Consequently, release 4.20 contains an out-of-date version. This
6807 shouldn't happen again.
6809 5. When building Exim with embedded Perl support, the script that builds the
6810 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6811 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6813 6. The freeze_tell option was not being used for messages that were frozen on
6814 arrival, either by an ACL or by local_scan().
6816 7. Added the smtp_incomplete_transaction log selector.
6818 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6819 was accepting AUTH without a new EHLO.
6821 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6822 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6825 10. Logging of TCP/IP connections (when configured) now happens in the main
6826 daemon process instead of the child process, so that the TCP/IP connection
6827 count is more accurate (but it can never be perfect).
6829 11. The use of "drop" in a nested ACL was not being handled correctly in the
6830 outer ACL. Now, if condition failure induced by the nested "drop" causes
6831 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6832 or "require"), the connection is dropped.
6834 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6835 that yield "discard" can now be used with an "accept" or a "discard" verb,
6836 but an error is generated for any others (because I can't see a useful way
6837 to define what should happen).
6839 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6840 are now processed in the same way as lines in the Exim configuration file.
6841 In particular, continuation lines are supported.
6843 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6845 15. Added -ti meaning -t -i.
6847 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6848 domains, and warn by logging if others are found.
6850 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6851 returned from GIFCONF when Exim is trying to find the list of interfaces on
6852 a host. The code in os.c has been modified to copy each ifreq to an aligned
6853 structure in all cases.
6855 Also, in some cases, the returned ifreq's were being copied to a 'struct
6856 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6857 means the last couple of bytes of an IPv6 address could be chopped if the
6858 ifreq contained only a normal sockaddr (14 bytes storage).
6860 18. Named domain lists were not supported in the hosts_treat_as_local option.
6861 An entry such as +xxxx was not recognized, and was treated as a literal
6864 19. Ensure that header lines added by a DATA ACL are included in the reject log
6865 if the ACL subsequently rejects the message.
6867 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6868 MD5 (which is deprecated).
6870 21. When testing a filter file using -bf, Exim was writing a message when it
6871 took the sender from a "From " line in the message, but it was not doing so
6872 when it took $return_path from a Return-Path: header line. It now does.
6874 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6875 with a valid header line field name (a series of printing characters
6876 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6878 23. Changed "disc" in the source to "disk" to conform to the documentation and
6879 the book and for uniformity.
6881 24. Ignore Sendmail's -Ooption=value command line item.
6883 25. When execve() failed while trying to run a command in a pipe transport,
6884 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6885 could be confused with a return value of 69 from the command itself. This
6886 has been changed to 127, the value the shell returns if it is asked to run
6887 a non-existent command. The wording for the related log line suggests a
6888 non-existent command as the problem.
6890 26. If received_header_text expands to an empty string, do not add a Received:
6891 header line to the message. (Well, it adds a token one on the spool, but
6892 marks it "old" so that it doesn't get used or transmitted.)
6894 27. Installed eximstats 1.28 (addition of -nt option).
6896 28. There was no check for failure on the call to getsockname() in the daemon
6897 code. This can fail if there is a shortage of resources on the system, with
6898 ENOMEM, for example. A temporary error is now given on failure.
6900 29. Contrary to the C standard, it seems that in some environments, the
6901 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6902 program. Exim now does this explicitly; it affects the formatting of
6903 timestamps using strftime().
6905 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6906 complaints. I've now initialized all the variables, to avoid this.
6908 32. Header lines added by a system filter were not being "seen" during
6909 transport-time rewrites.
6911 33. The info_callback() function passed to OpenSSL is set up with type void
6912 (*)(SSL *, int, int), as described somewhere. However, when calling the
6913 function (actually a macro) that sets it up, the type void(*)() is
6914 expected. I've put in a cast to prevent warnings from picky compilers.
6916 34. If a DNS black list lookup found a CNAME record, but there were no A
6917 records associated with the domain it pointed at, Exim crashed.
6919 35. If a DNS black list lookup returned more than one A record, Exim ignored
6920 all but the first. It now scans all returned addresses if a particular IP
6921 value is being sought. In this situation, the contents of the
6922 $dnslist_value variable are a list of all the addresses, separated by a
6925 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6926 to accept a host name and all its aliases if the forward lookup for any of
6927 them yielded the IP address of the incoming connection. Now it accepts only
6928 those names whose forward lookup yields the correct IP address. Any other
6929 names are discarded. This closes a loophole whereby a rogue DNS
6930 administrator could create reverse DNS records to break through a
6931 wildcarded host restriction in an ACL.
6933 37. If a user filter or a system filter that ran in a subprocess used any of
6934 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6935 the wrong values were passed to the pipe command ($thisaddress had the
6936 value of $0, $0 had the value of $1, etc). This bug was introduced by
6937 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6939 38. Improved the line breaking for long SMTP error messages from ACLs.
6940 Previously, if there was no break point between 40 and 75 characters, Exim
6941 left the rest of the message alone. Two changes have been made: (a) I've
6942 reduced the minimum length to 35 characters; (b) if it can't find a break
6943 point between 35 and 75 characters, it looks ahead and uses the first one
6944 that it finds. This may give the occasional overlong line, but at least the
6945 remaining text gets split now.
6947 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6948 file descriptors might be low, and that setting 1000 would always raise it.
6949 It turns out that in some environments, the limit is already over 1000 and
6950 that lowering it causes trouble. So now Exim takes care not to decrease it.
6952 40. When delivering a message, the value of $return_path is set to $sender_
6953 address at the start of routing (routers may change the value). By an
6954 oversight, this default was not being set up when an address was tested by
6955 -bt or -bv, which affected the outcome if any router or filter referred to
6958 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6959 log only when "message" or "log_message" are set. However, if one of the
6960 conditions was an address verification, or a call to a nested ACL, the
6961 messages generated by the underlying test were being passed through. This
6962 no longer happens. The underlying message is available in $acl_verify_
6963 message for both "message" and "log_message" expansions, so it can be
6964 passed through if needed.
6966 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6967 new expansion $bh_ to give the encoded byte string without charset
6968 translation. Translation happens only if iconv() is available; HAVE_ICONV
6969 indicates this at build time. HEADERS_CHARSET gives the charset to
6970 translate to; headers_charset can change it in the configuration, and
6971 "headers charset" can change it in an individual filter file.
6973 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6974 that creates RFC 2047 encoded "words" labels them as that charset instead
6975 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6976 expansion operator; (ii) when Exim creates a From: line for a local
6977 message; (iii) when a header line is rewritten to include a "phrase" part.
6979 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6980 buggy, causing it to skip the first lines of messages whose message ID
6981 ended in 'D'. This would not have bitten before Exim release 4.14, because
6982 message IDs were unlikely to end in 'D' before then. The effect was to have
6983 incorrect size information for certain domains.
6985 45. #include "config.h" was missing at the start of the crypt16.c module. This
6986 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6989 46. If there was a timeout during a "random" callout check, Exim treated it as
6990 a failure of the random address, and carried on sending RSET and the real
6991 address. If the delay was just some slowness somewhere, the response to the
6992 original RCPT would be taken as a response to RSET and so on, causing
6993 mayhem of various kinds.
6995 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6996 when I implemented it. It didn't allow for the fact that some option values
6997 may legitimately be negative (e.g. size_addition), and it didn't even do
6998 the right test for positive values.
7000 48. Domain names in DNS records are case-independent. Exim always looks them up
7001 in lower case. Some resolvers return domain names in exactly the case they
7002 appear in the zone file, that is, they may contain uppercase letters. Not
7003 all resolvers do this - some return always lower case. Exim was treating a
7004 change of case by a resolver as a change of domain, similar to a widening
7005 of a domain abbreviation. This triggered its re-routing code and so it was
7006 trying to route what was effectively the same domain again. This normally
7007 caused routing to fail (because the router wouldn't handle the domain
7008 twice). Now Exim checks for this case specially, and just changes the
7009 casing of the domain that it ultimately uses when it transmits the message
7012 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7015 50. If a filter generated a file delivery with a non-absolute name (possible if
7016 no home directory exists for the router), the forbid_file option was not
7019 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7020 the existing equality matching.
7022 52. Exim was using ints instead of ino_t variables in some places where it was
7023 dealing with inode numbers.
7025 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7026 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7027 called TMPDIR, and if it finds it is different, it changes its value.
7029 54. The smtp_printf() function is now made available to local_scan() so
7030 additional output lines can be written before returning. There is also an
7031 smtp_fflush() function to enable the detection of a dropped connection.
7032 The variables smtp_input and smtp_batched_input are exported to
7035 55. Changed the default runtime configuration: the message "Unknown user"
7036 has been removed from the ACL, and instead placed on the localuser router,
7037 using the cannot_route_message feature. This means that any verification
7038 failures that generate their own messages won't get overridden. Similarly,
7039 the "Unrouteable address" message that was in the ACL for unverifiable
7040 relay addresses has also been removed.
7042 56. Added hosts_avoid_esmtp to the smtp transport.
7044 57. The exicyclog script was not checking for the esoteric option
7045 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7046 will work only if exicyclog is run under the appropriate euid.
7048 58. Following a discussion on the list, the rules by which Exim recognises line
7049 endings on incoming messages have been changed. The -dropcr and drop_cr
7050 options are now no-ops, retained only for backwards compatibility. The
7051 following line terminators are recognized: LF CRLF CR. However, special
7052 processing applies to CR:
7054 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7055 nor a local message in the state where . is a terminator.
7057 (ii) If a bare CR is encountered in a header line, an extra space is added
7058 after the line terminator so as not to end the header. The reasoning
7059 behind this is that bare CRs in header lines are most likely either
7060 to be mistakes, or people trying to play silly games.
7062 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7063 was being incorrectly given as 18 bytes larger than it should have been.
7064 This is a VOB (very old bug).
7066 60. This may never have affected anything current, but just in case it has:
7067 When the local host is found other than at the start of a list of hosts,
7068 the local host, those with the same MX, and any that follow, are discarded.
7069 When the list in question was part of a longer list of hosts, the following
7070 hosts (not currently being processed) were also being discarded. This no
7071 longer happens. I'm not sure if this situation could ever has previously
7074 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7077 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7078 header. This is something that is recommended in a new Internet Draft, and
7079 is something that is documented as being done by Sendmail. There are two
7080 possible values. For messages generated by the autoreply transport, Exim
7083 Auto-Submitted: auto-replied
7085 whereas for all other generated messages (e.g. bounces) it adds
7087 Auto-Submitted: auto-generated
7089 63. The "personal" condition in filters now includes a test for the
7090 Auto-Submitted: header. If it contains the string "auto-" the message it
7091 not considered personal.
7093 64. Added rcpt_include_affixes as a generic transport option.
7095 65. Added queue_only_override (default true).
7097 66. Added the syslog_duplication option.
7099 67. If what should have been the first header line of a message consisted of
7100 a space followed by a colon, Exim was mis-interpreting it as a header line.
7101 It isn't of course - it is syntactically invalid and should therefore be
7102 treated as the start of the message body. The misbehaviour could have
7103 caused a number of strange effects, including loss of data in subsequent
7104 header lines, and spool format errors.
7106 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7107 client host had authenticated. This control can now be exercised by an ACL
7108 for more flexibility.
7110 69. By default, callouts do not happen when testing with -bh. There is now a
7111 variant, -bhc, which does actually run the callout code, including
7112 consulting and updating the callout cache.
7114 70. Added support for saslauthd authentication, courtesy of Alexander
7117 71. If statvfs() failed on the spool or log directories while checking their
7118 size for availability, Exim confusingly gave the error "space shortage".
7119 Furthermore, in debugging mode it crashed with a floating point exception.
7120 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7121 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7122 serious problem, Exim now writes to the main and panic logs when this
7123 happens, with details of the failure. It then refuses to accept the
7124 incoming message, giving the message "spool directory problem" or "log
7125 directory problem" with a 421 code for SMTP messages.
7127 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7128 0, 1, and 2 exist, because some OS complain for execs without them (see
7129 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7130 descriptors. However, the code omitted to check that the open succeeded,
7131 causing mysterious errors if for some reason the permissions on /dev/null
7132 got screwed. Now Exim writes a message to the main and panic logs, and
7133 bombs out if it can't open /dev/null.
7135 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7136 interact so that it is all more flexible. It is supposed to remain
7137 backwards compatible. Also added extra_local_interfaces.
7139 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7140 to bomb out with an assertion failure - to the client this appears as a
7141 connection drop. This problem occurs in the part of the code that was taken
7142 from the Samba project. Fortunately, the assertion is in a very simple
7143 function, so I have fixed this by reproducing the function inline in the
7144 one place where it is called, and arranging for authentication to fail
7145 instead of killing the process with assert().
7147 75. The SPA client code was not working when the server requested OEM rather
7148 than Unicode encoding.
7150 76. Added code to make require_files with a specific uid setting more usable in
7151 the case where statting the file as root fails - usually a non-root-mounted
7152 NFS file system. When this happens and the failure is EACCES, Exim now
7153 forks a subprocess and does the per-uid checking as the relevant uid.
7155 77. Added process_log_path.
7157 78. If log_file_path was not explicitly set, a setting of check_log_space or
7158 check_log_inodes was ignored.
7160 79. If a space check for the spool or log partitions fails, the incident is now
7161 logged. Of course, in the latter case the data may get lost...
7163 80. Added the %p formatting code to string_format() so that it can be used to
7164 print addresses in debug_print(). Adjusted all the address printing in the
7165 debugging in store.c to use %p rather than %d.
7167 81. There was a concern that a line of code in smtp_in.c could overflow a
7168 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7169 initially expressed, the concern was not well-founded, because trailing
7170 spaces are removed early. However, if the trailing spaces were followed by
7171 a NULL, they did not get removed, so the overflow was possible. Two fixes
7174 (a) I re-wrote the offending code in a cleaner fashion.
7175 (b) If an incoming SMTP command contains a NULL character, it is rejected
7178 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7179 runs initgroups(), so that if the Exim user is in any additional groups,
7180 they will be used during message reception.
7186 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob