1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
29 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
30 When redis_servers is set to a list of > 1 element, and the Redis servers
31 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
32 case of MOVED into a DEFER case instead, thus moving the query onto the
33 next server in the list. For a cluster of N elements, all N servers must
34 be defined in redis_servers.
36 GF/02 Catch and remove uninitialized value warning in exiqsumm
37 Check for existence of @ARGV before looking at $ARGV[0]
39 JH/01 Replace the store_release() internal interface with store_newblock(),
40 which internalises the check required to safely use the old one, plus
41 the allocate and data copy operations duplicated in both (!) of the
44 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
45 modifier. This matches the restriction on the commandline.
47 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
48 Previously only the last row was returned.
50 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
51 we assumed that tags in the header were well-formed, and parsed the
52 element content after inspecting only the first char of the tag.
53 Assumptions at that stage could crash the receive process on malformed
56 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
57 While running the DKIM ACL we operate on the Permanent memory pool so that
58 variables created with "set" persist to the DATA ACL. Also (at any time)
59 DNS lookups that fail create cache records using the Permanent pool. But
60 expansions release any allocations made on the current pool - so a dnsdb
61 lookup expansion done in the DKIM ACL releases the memory used for the
62 DNS negative-cache, and bad things result. Solution is to switch to the
63 Main pool for expansions.
64 While we're in that code, add checks on the DNS cache during store_reset,
65 active in the testsuite.
66 Problem spotted, and debugging aided, by Wolfgang Breyha.
68 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
69 When none of the hosts presented to a transport match an already-open
70 connection, close it and proceed with the list. Previously we would
71 queue the message. Spotted by Lena with Yahoo, probably involving
74 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
75 Previously a spurious "250 OK id=" response was appended to the proper
78 JH/08 The "support for" informational output now, which built with Content
79 Scanning support, has a line for the malware scanner interfaces compiled
80 in. Interface can be individually included or not at build time.
82 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
83 by the template makefile "src/EDITME". The "STREAM" support for an older
84 ClamAV interface method is removed.
86 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
87 rows affected is given instead).
89 JH/11 The runtime Berkeley DB library version is now additionally output by
90 "exim -d -bV". Previously only the compile-time version was shown.
92 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
93 SMTP connection. Previously, when one had more receipients than the
94 first, an abortive onward connection was made. Move to full support for
95 multiple onward connections in sequence, handling cutthrough connection
96 for all multi-message initiating connections.
98 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
99 routers. Previously, a multi-recipient message would fail to match the
100 onward-connection opened for the first recipient, and cause its closure.
102 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
103 a timeout on read on a GnuTLS initiating connection, resulting in the
104 initiating connection being dropped. This mattered most when the callout
105 was marked defer_ok. Fix to keep the two timeout-detection methods
108 JH/15 Relax results from ACL control request to enable cutthrough, in
109 unsupported situations, from error to silently (except under debug)
110 ignoring. This covers use with PRDR, frozen messages, queue-only and
113 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
115 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
116 metadata, resulting in a crash in free().
118 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
119 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
120 Broken also in d185889f4, with init system revamp.
122 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
123 Previously we abruptly closed the connection after reading a malware-
124 found indication; now we go on to read the "scan ok" response line,
127 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
128 ACL. Previously, a crash would result.
130 JH/19 Speed up macro lookups during configuration file read, by skipping non-
131 macro text after a replacement (previously it was only once per line) and
132 by skipping builtin macros when searching for an uppercase lead character.
134 JH/20 DANE support moved from Experimental to mainline. The Makefile control
135 for the build is renamed.
137 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
138 was allocated for every new TLS startup, meaning one per message. Fix
139 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
141 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
142 reported the original. Fix to report (as far as possible) the ACL
143 result replacing the original.
145 JH/23 Fix memory leak during multi-message connections using STARTTLS under
146 OpenSSL. Certificate information is loaded for every new TLS startup,
147 and the resources needed to be freed.
149 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
151 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
154 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
155 DATA response info to the (existing) per-recipient response info for
156 the "C=" log element. It can have useful tracking info from the
157 destination system. Patch from Simon Arlott.
159 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
160 length value. Previously this would segfault.
162 HS/02 Support Avast multiline protoocol, this allows passing flags to
163 newer versions of the scanner.
165 JH/28 Ensure that variables possibly set during message acceptance are marked
166 dead before release of memory in the daemon loop. This stops complaints
167 about them when the debug_store option is enabled. Discovered specifically
168 for sender_rate_period, but applies to a whole set of variables.
169 Do the same for the queue-runner and queue-list loops, for variables set
170 from spool message files. Do the same for the SMTP per-message loop, for
171 certain variables indirectly set in ACL operations.
173 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
174 as a multi-recipient message from a mailinglist manager). The coding had
175 an arbitrary cutoff number of characters while checking for more input;
176 enforced by writing a NUL into the buffer. This corrupted long / fast
177 input. The problem was exposed more widely when more pipelineing of SMTP
178 responses was introduced, and one Exim system was feeding another.
179 The symptom is log complaints of SMTP syntax error (NUL chars) on the
180 receiving system, and refused recipients seen by the sending system
181 (propating to people being dropped from mailing lists).
182 Discovered and pinpointed by David Carter.
184 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
185 replaced by the ${authresults } expansion.
187 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
189 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
190 allows proper process termination in container environments.
192 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
193 Previously the "final dot" had a newline after it; ensure it is CR,LF.
195 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
196 and "err_perm", deprecated since 4.83 when the RFC-defined words
197 "temperror" and "permerror" were introduced.
199 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
200 transport-filters or DKIM-signing. The restriction was lost in the
201 consolidation of verify-callout and delivery SMTP handling.
202 Extend the restriction to also cover ARC-signing.
204 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
205 in defer=pass mode supply a 450 to the initiator. Previously the message
208 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
209 tls_require_ciphers is used as before.
211 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
212 "pass_unscanned". Only tmpfails from the scanner are written to
213 the paniclog, as they may require admin intervention (permission
214 denied, license issues). Other scanner errors (like decompression
215 bombs) do not cause a paniclog entry.
217 JH/36 Fix reinitialisation of DKIM logging variable between messages.
218 Previously it was possible to log spurious information in receive log
221 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
222 triggered odd behaviour from Outlook Express clients.
224 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
227 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
228 since the IETF WG has not yet settled on that versus the original
229 "bare" representation.
231 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
232 Previously the millisecond value corrupted the output.
233 Fix also for syslog_pid=no and log_selector +pid, for which the pid
234 corrupted the output.
240 JH/01 Rework error string handling in TLS interface so that the caller in
241 more cases is responsible for logging. This permits library-sourced
242 string to be attached to addresses during delivery, and collapses
243 pairs of long lines into single ones.
245 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
246 during configuration. Wildcards are allowed and expanded.
248 JH/02 Rework error string handling in DKIM to pass more info back to callers.
249 This permits better logging.
251 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
252 do not close it down and have the child transport start it up again on
253 the passed-on TCP connection. Instead, proxy the child (and any
254 subsequent ones) for TLS via a unix-domain socket channel. Logging is
255 affected: the continued delivery log lines do not have any DNSSEC, TLS
256 Certificate or OCSP information. TLS cipher information is still logged.
258 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
259 identical IP addresses on different listening ports. Will also affect
262 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
263 add noisy ifdef guards to special-case this sillyness.
264 Patch from Bernd Kuhls.
266 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
267 than 255 are no longer allowed.
269 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
270 Disable the session-cache too, which might reduce our load. Since we
271 currrectly use a new context for every connection, both as server and
272 client, there is no benefit for these.
273 GnuTLS appears to not support tickets server-side by default (we don't
274 call gnutls_session_ticket_enable_server()) but client side is enabled
275 by default on recent versions (3.1.3 +) unless the PFS priority string
278 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
279 <https://reproducible-builds.org/specs/source-date-epoch/>.
281 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
282 the check for any unsuccessful recipients did not notice the limit, and
283 erroneously found still-pending ones.
285 JH/08 Pipeline CHUNKING command and data together, on kernels that support
286 MSG_MORE. Only in-clear (not on TLS connections).
288 JH/09 Avoid using a temporary file during transport using dkim. Unless a
289 transport-filter is involved we can buffer the headers in memory for
290 creating the signature, and read the spool data file once for the
291 signature and again for transmission.
293 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
294 4.77 as the kernel support then wasn't solid, having issues in 64bit
295 mode. Now, it's been long enough. Add support for FreeBSD also.
297 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
298 case where the routing stage had gathered several addresses to send to
299 a host before calling the transport for the first, we previously failed
300 to close down TLS in the old transport process before passing the TCP
301 connection to the new process. The new one sent a STARTTLS command
302 which naturally failed, giving a failed delivery and bloating the retry
303 database. Investigation and fix prototype from Wolfgang Breyha.
305 JH/12 Fix check on SMTP command input synchronisation. Previously there were
306 false-negatives in the check that the sender had not preempted a response
307 or prompt from Exim (running as a server), due to that code's lack of
308 awareness of the SMTP input buffering.
310 PP/04 Add commandline_checks_require_admin option.
311 Exim drops privileges sanely, various checks such as -be aren't a
312 security problem, as long as you trust local users with access to their
313 own account. When invoked by services which pass untrusted data to
314 Exim, this might be an issue. Set this option in main configuration
315 AND make fixes to the calling application, such as using `--` to stop
318 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
319 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
320 responses to those, into a single TLS record each way (this usually means
321 a single packet). As a side issue, smtp_enforce_sync now works on TLS
324 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
325 affects you only if you're dancing at the edge of the param size limits.
326 If you are, and this message makes sense to you, then: raise the
327 configured limit or use OpenSSL 1.1. Nothing we can do for older
330 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
331 cmdline element to get the documented default one. Previously it was
334 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
335 are unsuable for non-message-specific reasons - eg. connection timeout,
336 banner-time rejection.
338 JH/16 Fix logging of delivery remote port, when specified by router, under
341 PP/06 Repair manualroute's ability to take options in any order, even if one
342 is the name of a transport.
345 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
347 JH/17 Change the list-building routines interface to use the expanding-string
348 triplet model, for better allocation and copying behaviour.
350 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
351 Previously it was constructed the first time a possibly-matching string
352 was met in the configuration file input during startup; now it is done
355 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
356 compatible one, to avoid the (poorly documented) possibility of a config
357 file in the working directory redirecting the DB files, possibly correpting
358 some existing file. CVE-2017-10140 assigned for BDB.
360 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
361 cache-hot. Previously, although the result was properly cached, the
362 initial verify call returned a defer.
364 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
365 the main verify for receipient in uncached-mode.
367 JH/22 Retire historical build files to an "unsupported" subdir. These are
368 defined as "ones for which we have no current evidence of testing".
370 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
371 if present. Previously it was ignored.
373 JH/24 Start using specified-initialisers in C structure init coding. This is
374 a C99 feature (it's 2017, so now considered safe).
376 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
377 if was a fixed-sized field and bitmask ops via macros; it is now more
380 PP/07 GitHub PR 56: Apply MariaDB build fix.
381 Patch provided by Jaroslav Škarvada.
383 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
384 during Coverity cleanups [4.87 JH/47]
385 Diagnosis and fix provided by Michael Fischer v. Mollard.
387 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
388 the right size to place the terminating semicolon on its own folded
389 line, the header hash was calculated to an incorrect value thanks to
390 the (relaxed) space the fold became.
392 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
393 and confused the parent.
395 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
396 which could crash as a result. This could lead to undeliverable messages.
398 JH/28 Logging: "next input sent too soon" now shows where input was truncated
401 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
402 matters on fast-turnover and PID-randomising systems, which were getting
403 out-of-order delivery.
405 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
406 a possibly-overlapping copy. The symptom was that "Remote host closed
407 connection in response to HELO" was logged instead of the actual 4xx
410 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
411 Previously only that bufferd was discarded, resulting in SYMTP command
414 JH/32 DKIM: when a message has multiple signatures matching an identity given
415 in dkim_verify_signers, run the dkim acl once for each. Previously only
416 one run was done. Bug 2189.
418 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
419 "panic the current process" to "deliberately defer". The panic log is
420 still written with the problem list name; the mail and reject logs now
421 get a temp-reject line for the message that was being handled, saying
422 something like "domains check lookup or other defer". The SMTP 451
423 message is still "Temporary local problem".
425 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
426 A crafted sequence of BDAT commands could result in in-use memory beeing
427 freed. CVE-2017-16943.
429 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
430 from SMTP input. Previously it was always done; now only done for DATA
431 and not BDAT commands. CVE-2017-16944.
433 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
434 to the message (such as an overlong header line). Previously this was
435 not done and we did not exit BDAT mode. Followon from the previous item
436 though a different problem.
442 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
443 than -2003 did; needs libidn2 in addition to libidn.
445 JH/02 The path option on a pipe transport is now expanded before use.
447 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
448 Patch provided by "Björn", documentation fix added too.
450 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
451 missing a wire-to-host endian conversion.
453 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
454 close after a BDAT command line could be taken as a following command,
455 giving a synch failure. Fix by only checking for synch immediately
456 before acknowledging the chunk.
458 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
459 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
460 macro. Patches provided by Josh Soref.
462 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
463 Previously we did not; the RFC seems ambiguous and VRFY is not listed
464 by IANA as a service extension. However, John Klensin suggests that we
467 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
468 the dkim code may be unix-mode line endings rather than smtp wire-format
469 CRLF, so prepend a CR to any bare LF.
471 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
472 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
474 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
475 the first were themselves being wrongly included in the feed into dkim
476 processing; with most chunk sizes in use this resulted in an incorrect
477 body hash calculated value.
479 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
480 DKIM signature block, for verification. Although advised against by
481 standards it is specifically not ruled illegal.
483 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
485 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
486 missing a body hash (the bh= tag).
488 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
489 It seems that HAProxy sends the Proxy Protocol information in clear and
490 only then does a TLS startup, so do the same.
492 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
493 TCP connections (such as for Spamd) unless the daemon successfully set
494 Fast Open mode on its listening sockets. This fixes breakage seen on
495 too-old kernels or those not configured for Fast Open, at the cost of
496 requiring both directions being enabled for TFO, and TFO never being used
497 by non-daemon-related Exim processes.
499 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
500 endings, at least on the first header line. Try to canonify any that get
501 past that check, despite the cost.
503 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
504 now limited to an arbitrary five deep, while parsing addresses with the
505 strip_excess_angle_brackets option enabled.
507 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
508 instead leave the unprompted TLS handshake in socket buffer for the
509 TLS library to consume.
511 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
513 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
515 JH/16 Drop variables when they go out of scope. Memory management drops a whole
516 region in one operation, for speed, and this leaves assigned pointers
517 dangling. Add checks run only under the testsuite which checks all
518 variables at a store-reset and panics on a dangling pointer; add code
519 explicitly nulling out all the variables discovered. Fixes one known
520 bug: a transport crash, where a dangling pointer for $sending_ip_address
521 originally assigned in a verify callout, is re-used.
523 PP/06 Drop '.' from @INC in various Perl scripts.
525 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
527 PP/08 Reduce a number of compilation warnings under clang; building with
528 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
529 should be warning-free.
531 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
533 HS/01 Fix portability problems introduced by PP/08 for platforms where
534 realloc(NULL) is not equivalent to malloc() [SunOS et al].
536 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
537 chunk. This allows us to accept broken chunked messages. We need a more
538 general solution here.
540 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
541 already-broken messages in the queue.
543 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
545 JH/19 Fix reference counting bug in routing-generated-address tracking.
551 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
552 supports it and a size is available (ie. the sending peer gave us one).
554 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
555 years of being deprecated). The replacements are the ACLs
556 acl_smtp_mime and acl_not_smtp_mime.
558 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
559 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
560 or even an in-clear connection were permitted. Now, if the host lookup
561 was dnssec and dane was requested then the host is only used if the
562 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
563 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
564 if one fails this test.
565 This means that a poorly-configured remote DNS will make it incommunicado;
566 but it protects against a DNS-interception attack on it.
568 JH/04 Bug 1810: make continued-use of an open smtp transport connection
569 non-noisy when a race steals the message being considered.
571 JH/05 If main configuration option tls_certificate is unset, generate a
572 self-signed certificate for inbound TLS connections.
574 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
575 in rewrites and routers.
577 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
578 and logged a warning sing 4.83; now they are a configuration file error.
580 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
581 (lacking @domain). Apply the same qualification processing as RCPT.
583 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
585 JH/10 Support ${sha256:} applied to a string (as well as the previous
588 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
589 a cutthrough deliver is pending, as we always want to make a connection.
590 This also avoids re-routing the message when later placing the cutthrough
591 connection after a verify cache hit.
592 Do not update it with the verify result either.
594 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
595 when routing results in more than one destination address.
597 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
598 signing (which inhibits the cutthrough capability). Previously only
599 the presence of an option was tested; now an expansion evaluating as
600 empty is permissible (obviously it should depend only on data available
601 when the cutthrough connection is made).
603 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
604 the relevant preceding SMTP command did not note the pipelining mode.
606 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
607 Previously they were not counted.
609 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
610 as one having no matching records. Previously we deferred the message
611 that needed the lookup.
613 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
614 distinguished as "(=".
616 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
617 for missing MX records. Previously it only worked for missing A records.
619 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
621 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
622 after the data-go-ahead and data-ack. Patch from Jason Betts.
624 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
625 even for a "none" policy. Patch from Tony Meyer.
627 JH/22 Fix continued use of a connection for further deliveries. If a port was
628 specified by a router, it must also match for the delivery to be
631 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
632 When one of the recipients of a message was unsuitable for the connection
633 (has no matching addresses), we lost track of needing to mark it
634 deferred. As a result mail would be lost.
636 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
638 JH/25 Decoding ACL controls is now done using a binary search; the source code
639 takes up less space and should be simpler to maintain. Merge the ACL
640 condition decode tables also, with similar effect.
642 JH/26 Fix problem with one_time used on a redirect router which returned the
643 parent address unchanged. A retry would see the parent address marked as
644 delivered, so not attempt the (identical) child. As a result mail would
647 JH/27 Fix a possible security hole, wherein a process operating with the Exim
648 UID can gain a root shell. Credit to http://www.halfdog.net/ for
649 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
652 JH/28 Enable {spool,log} filesystem space and inode checks as default.
653 Main config options check_{log,spool}_{inodes,space} are now
654 100 inodes, 10MB unless set otherwise in the configuration.
656 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
657 Previously it only applied to the main-section connection policy
660 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
662 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
663 by me. Added RFC7919 DH primes as an alternative.
665 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
666 are not in the system include path.
668 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
669 GnuTLS, when a session startup failed (eg because the client disconnected)
670 Exim did stdio operations after fclose. This was exposed by a recent
671 change which nulled out the file handle after the fclose.
673 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
674 signed directly by the cert-signing cert, rather than an intermediate
675 OCSP-signing cert. This is the model used by LetsEncrypt.
677 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
679 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
680 an incoming connection.
682 HS/02 Bug 1802: Do not half-close the connection after sending a request
685 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
686 fallback to "prime256v1".
688 JH/34 SECURITY: Use proper copy of DATA command in error message.
689 Could leak key material. Remotely exploitable. CVE-2016-9963.
695 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
696 and 3.4.4 - once the server is enabled to respond to an OCSP request
697 it does even when not requested, resulting in a stapling non-aware
698 client dropping the TLS connection.
700 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
701 support variable-length bit vectors. No functional change.
703 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
704 The I= interface field on outgoing lines is now after the H= remote
705 host field, same as incoming lines. There is a separate
706 outgoing_interface log selector which allows you to disable the
709 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
710 If not running log_selector +smtp_connection the mainlog would be held
711 open indefinitely after a "too many connections" event, including to a
712 deleted file after a log rotate. Leave the per net connection logging
713 leaving it open for efficiency as that will be quickly detected by the
714 check on the next write.
716 HS/01 Bug 1671: Fix post transport crash.
717 Processing the wait-<transport> messages could crash the delivery
718 process if the message IDs didn't exist for some reason. When
719 using 'split_spool_directory=yes' the construction of the spool
720 file name failed already, exposing the same netto behaviour.
722 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
723 mime_regex ACL conditions.
725 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
726 to DSN fail messages (bounces): remote IP, remote greeting, remote response
727 to HELO, local diagnostic string.
729 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
730 log line to debug. Even when configured with a tls authenticator many
731 client connections are expected to not authenticate in this way, so
732 an authenticate fail is not an error.
734 HS/02 Add the Exim version string to the process info. This way exiwhat
735 gives some more detail about the running daemon.
737 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
738 matter for fast-change records such as DNSBLs.
740 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
741 retry record, even if constant. There may be multiple transports with
742 different interface settings and the retry behaviour needs to be kept
745 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
747 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
749 JH/10 Bug 840: fix log_defer_output option of pipe transport
751 JH/11 Bug 830: use same host for all RCPTS of a message, even under
752 hosts_randomize. This matters a lot when combined with mua_wrapper.
754 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
755 ${quote_pgsql:<string>} operator.
757 JH/13 Bug 1708: avoid misaligned access in cached lookup.
759 JH/14 Change header file name for freeradius-client. Relevant if compiling
760 with Radius support; from the Gentoo tree and checked under Fedora.
762 JH/15 Bug 1712: Introduce $prdr_requested flag variable
764 JH/16 Bug 1714: Permit an empty string as expansion result for transport
765 option transport_filter, meaning no filtering.
767 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
769 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
770 defaults to "*" (all hosts). The variable is now available when not built
771 with TLS, default unset, mainly to enable keeping the testsuite sane.
772 If a server certificate is not supplied (via tls_certificate) an error is
773 logged, and clients will find TLS connections fail on startup. Presumably
774 they will retry in-clear.
775 Packagers of Exim are strongly encouraged to create a server certificate
776 at installation time.
778 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
779 with the $config_file variable.
781 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
782 in transport context, after the attempt, and per-recipient. The latter type
783 is per host attempted. The event data is the error message, and the errno
784 information encodes the lookup type (A vs. MX) used for the (first) host,
785 and the trailing two digits of the smtp 4xx response.
787 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
788 to write to mainlog (or rejectlog, paniclog) in the window between file
789 creation and permissions/ownership being changed. Particularly affects
790 installations where exicyclog is run as root, rather than exim user;
791 result is that the running daemon panics and dies.
793 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
795 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
796 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
797 "pri" and "weight". Note that the previous implicit priority given by the
798 list order is no longer honoured.
800 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
803 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
804 by defining SUPPORT_SOCKS.
806 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
807 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
808 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
809 variables are renamed to proxy_{local,external}_{address,port}.
811 JH/27 Move Internationalisation support from Experimental to mainline, enabled
812 for a build by defining SUPPORT_I18N
814 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
815 of the query string, and make ${quote_redis:} do that quoting.
817 JH/29 Move Events support from Experimental to mainline, enabled by default
818 and removable for a build by defining DISABLE_EVENT.
820 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
822 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
823 cached by the daemon.
825 JH/32 Move Redis support from Experimental to mainline, enabled for a build
826 by defining LOOKUP_REDIS. The libhiredis library is required.
828 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
829 keys are given for lookup.
831 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
832 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
833 only supported when built with TLS support. The PolarSSL SHA routines
834 are still used when the TLS library is too old for convenient support.
836 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
837 openssl_options), for security. OpenSSL forces this from version 1.1.0
838 server-side so match that on older versions.
840 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
841 allocation for $value could be released as the expansion processing
842 concluded, but leaving the global pointer active for it.
844 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
845 and to use the domains and local_parts ACL conditions.
847 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
848 incorrectly not doubled on cutthrough transmission, hence seen as a
849 body-termination at the receiving system - resulting in truncated mails.
850 Commonly the sender saw a TCP-level error, and retransmitted the message
851 via the normal store-and-forward channel. This could result in duplicates
852 received - but deduplicating mailstores were liable to retain only the
853 initial truncated version.
855 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
857 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
859 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
860 we're in there, support oversigning also; bug 1309.
862 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
864 HS/04 Add support for keep_environment and add_environment options.
866 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
867 either intentional arithmetic overflow during PRNG, or testing config-
870 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
871 delivery resulted in actual delivery. Cancel cutthrough before DATA
874 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
875 rejects a recipient: pass the reject to the originator.
877 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
878 Many were false-positives and ignorable, but it's worth fixing the
881 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
882 for the new environment-manipulation done at startup. Move the routines
883 from being local to tls.c to being global via the os.c file.
885 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
886 an extract embedded as result-arg for a map, the first arg for extract
887 is unavailable so we cannot tell if this is a numbered or keyed
888 extraction. Accept either.
894 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
897 JH/02 The smtp transport option "multi_domain" is now expanded.
899 JH/03 The smtp transport now requests PRDR by default, if the server offers
902 JH/04 Certificate name checking on server certificates, when exim is a client,
903 is now done by default. The transport option tls_verify_cert_hostnames
904 can be used to disable this per-host. The build option
905 EXPERIMENTAL_CERTNAMES is withdrawn.
907 JH/05 The value of the tls_verify_certificates smtp transport and main options
908 default to the word "system" to access the system default CA bundle.
909 For GnuTLS, only version 3.0.20 or later.
911 JH/06 Verification of the server certificate for a TLS connection is now tried
912 (but not required) by default. The verification status is now logged by
913 default, for both outbound TLS and client-certificate supplying inbound
916 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
919 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
920 Status Notification (bounce) messages are now MIME format per RFC 3464.
921 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
922 under the control of the dsn_advertise_hosts option, and routers may
923 have a dsn_lasthop option.
925 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
926 default, modifiable by a malware= option. The list separator for
927 the options can now be changed in the usual way. Bug 68.
929 JH/10 The smtp_receive_timeout main option is now expanded before use.
931 JH/11 The incoming_interface log option now also enables logging of the
932 local interface on delivery outgoing connections.
934 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
935 if the interface and destination host and port all match.
937 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
940 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
941 Patch from Andrew Lewis.
943 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
944 now supports optional time-restrictions, weighting, and priority
945 modifiers per server. Patch originally by <rommer@active.by>.
947 JH/16 The spamd_address main option now supports a mixed list of local
948 and remote servers. Remote servers can be IPv6 addresses, and
949 specify a port-range.
951 JH/17 Bug 68: The spamd_address main option now supports an optional
952 timeout value per server.
954 JH/18 Bug 1581: Router and transport options headers_add/remove can
955 now have the list separator specified.
957 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
960 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
963 JH/21 Support for the A6 type of dns record is withdrawn.
965 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
966 rather than the verbs used.
968 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
969 from 255 to 1024 chars.
971 JH/24 Verification callouts now attempt to use TLS by default.
973 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
974 are generic router options now. The defaults didn't change.
976 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
977 Original patch from Alexander Shikoff, worked over by JH.
979 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
980 files if ZCAT_COMMAND is not executable.
982 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
984 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
986 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
987 Normally benign, it bites when the pair was led to by a CNAME;
988 modern usage is to not canonicalize the domain to a CNAME target
989 (and we were inconsistent anyway for A-only vs AAAA+A).
991 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
993 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
994 when evaluating $sender_host_dnssec.
996 JH/31 Check the HELO verification lookup for DNSSEC, adding new
997 $sender_helo_dnssec variable.
999 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1001 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1003 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1005 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1006 documented as working, but never had. Support all but $spam_report.
1008 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1009 added for tls authenticator.
1011 HS/03 Add perl_taintmode main config option
1017 TL/01 When running the test suite, the README says that variables such as
1018 no_msglog_check are global and can be placed anywhere in a specific
1019 test's script, however it was observed that placement needed to be near
1020 the beginning for it to behave that way. Changed the runtest perl
1021 script to read through the entire script once to detect and set these
1022 variables, reset to the beginning of the script, and then run through
1023 the script parsing/test process like normal.
1025 TL/02 The BSD's have an arc4random API. One of the functions to induce
1026 adding randomness was arc4random_stir(), but it has been removed in
1027 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1028 function when detected.
1030 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1031 cause callback expansion.
1033 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1034 syntax errors in an expansion can be treated as a string instead of
1035 logging or causing an error, due to the internal use of bool_lax
1036 instead of bool when processing it.
1038 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1039 server certificates when making smtp deliveries.
1041 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1043 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1045 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1047 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1048 Merged patch from Sebastian Wiedenroth.
1050 JH/05 Fix results-pipe from transport process. Several recipients, combined
1051 with certificate use, exposed issues where response data items split
1052 over buffer boundaries were not parsed properly. This eventually
1053 resulted in duplicates being sent. This issue only became common enough
1054 to notice due to the introduction of connection certificate information,
1055 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1057 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1058 size buffer was used, resulting in syntax errors when an expansion
1061 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1062 version 3.3.6 or later.
1064 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1065 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1066 both become "event_action", the variables become $event_name, $event_data
1067 and $event_defer_errno. There is a new variable $verify_mode, usable in
1068 routers, transports and related events. The tls:cert event is now also
1069 raised for inbound connections, if the main configuration event_action
1072 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1073 early OCSP support, but no stapling (appears to be less than 1.0.0).
1075 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1076 server certificate names available under the smtp transport option
1077 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1080 JH/10 Time-related extraction expansions from certificates now use the main
1081 option "timezone" setting for output formatting, and are consistent
1082 between OpenSSL and GnuTLS compilations. Bug 1541.
1084 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1085 encoded parameter in the incoming message. Bug 1558.
1087 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1088 include certificate info, eximon was claiming there were spoolfile
1091 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1093 JH/14 Log delivery-related information more consistently, using the sequence
1094 "H=<name> [<ip>]" wherever possible.
1096 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1097 are problematic for Debian distribution, omit them from the release
1100 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1102 JH/16 Fix string representation of time values on 64bit time_t architectures.
1105 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1106 output list separator was used.
1111 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1112 checkers that were complaining about end of non-void function with no
1115 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1116 This was a regression introduced in 4.83 by another bugfix.
1118 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1120 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1121 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1127 TF/01 Correctly close the server side of TLS when forking for delivery.
1129 When a message was received over SMTP with TLS, Exim failed to clear up
1130 the incoming connection properly after forking off the child process to
1131 deliver the message. In some situations the subsequent outgoing
1132 delivery connection happened to have the same fd number as the incoming
1133 connection previously had. Exim would try to use TLS and fail, logging
1134 a "Bad file descriptor" error.
1136 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1137 utilities have not been installed.
1139 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1140 temporary space as the ACL may create new global variables.
1142 TL/01 LDAP support uses per connection or global context settings, depending
1143 upon the detected version of the libraries at build time.
1145 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1146 to extract and use the src ip:port in logging and expansions as if it
1147 were a direct connection from the outside internet. PPv2 support was
1148 updated based on HAProxy spec change in May 2014.
1150 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1152 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1153 Properly escape header and check for NULL return.
1155 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1158 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1160 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1161 characters in header names, implemented as a verify condition.
1162 Contributed by Michael Fischer v. Mollard.
1164 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1165 results permerror and temperror. Previous values are deprecated but
1166 still accepted. In a future release, err_perm and err_temp will be
1167 completely removed, which will be a backward incompatibility if the
1168 ACL tests for either of these two old results. Patch contributed by
1169 user bes-internal on the mailing list.
1171 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1173 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1174 selectors, in both main and reject logs.
1176 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1179 JH/07 Add malware type "sock" for talking to simple daemon.
1181 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1183 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1184 routers/transports under cutthrough routing.
1186 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1187 numbers. Touch up "bool" conditional to keep the same definition.
1189 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1191 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1193 JH/12 Expand items in router/transport headers_add or headers_remove lists
1194 individually rather than the list as a whole. Bug 1452.
1196 Required for reasonable handling of multiple headers_ options when
1197 they may be empty; requires that headers_remove items with embedded
1198 colons must have them doubled (or the list-separator changed).
1200 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1201 view the policy declared in the DMARC record. Currently, $dmarc_status
1202 is a combined value of both the record presence and the result of the
1205 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1207 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1208 dnslookup router and the smtp transport (applying to the forward
1211 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1212 of ldap servers used for a specific lookup. Patch provided by Heiko
1215 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1216 New variable $lookup_dnssec_authenticated for observability.
1218 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1219 Patch submitted by Lars Timman.
1221 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1223 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1224 Requires trusted mode and valid format message id, aborts otherwise.
1225 Patch contributed by Heiko Schlichting.
1227 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1228 certextract with support for various fields. Bug 1358.
1230 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1231 is requested by default, modifiable by smtp transport option
1234 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1235 operate on certificate variables to give certificate fingerprints
1236 Also new ${sha256:cert_variable}.
1238 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1240 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1241 Christian Aistleitner.
1243 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1245 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1246 file. Patch from Wolfgang Breyha.
1248 JH/25 Expand the coverage of the delivery $host and $host_address to
1249 client authenticators run in verify callout. Bug 1476.
1251 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1252 align with daemon_smtp_ports. Bug 72.
1254 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1255 support and error reporting did not work properly.
1257 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1258 and is readable. Patch from Andrew Colin Kissa.
1260 TL/14 Enhance documentation of ${run expansion and how it parses the
1261 commandline after expansion, particularly in the case when an
1262 unquoted variable expansion results in an empty value.
1264 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1266 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1267 by renaming away from it. Addresses GH issue 15, reported by
1270 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1272 TL/15 SECURITY: prevent double expansion in math comparison functions
1273 (can expand unsanitized data). Not remotely exploitable.
1280 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1282 PP/02 Make -n do something, by making it not do something.
1283 When combined with -bP, the name of an option is not output.
1285 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1288 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1289 $sender_host_name and config options to manage this, and basic check
1292 PP/05 DSCP support for outbound connections and control modifier for inbound.
1294 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1295 (Only plugin which currently uses this is kerberos4, which nobody should
1296 be using, but we should make it available and other future plugins might
1297 conceivably use it, even though it would break NAT; stuff *should* be
1298 using channel bindings instead).
1300 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1301 name; added for Sendmail compatibility; requires admin caller.
1302 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1303 just ignore it); requires trusted caller.
1304 Also parse but ignore: -Ac -Am -X<logfile>
1307 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1309 TL/02 Add +smtp_confirmation as a default logging option.
1311 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1312 Patch by Magnus Holmgren from 2007-02-20.
1314 TL/04 Bugzilla 1281 - Spec typo.
1315 Bugzilla 1283 - Spec typo.
1316 Bugzilla 1290 - Spec grammar fixes.
1318 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1320 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1322 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1323 mailing list by by Dmitry Isaikin.
1325 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1327 JH/02 Support "G" suffix to numbers in ${if comparisons.
1329 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1331 NM/01 Bugzilla 1197 - Spec typo
1332 Bugzilla 1196 - Spec examples corrections
1334 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1336 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1337 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1340 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1341 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1343 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1344 "acl {{name}{arg}...}", and optional args on acl condition
1347 JH/05 Permit multiple router/transport headers_add/remove lines.
1349 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1351 JH/07 Avoid using a waiting database for a single-message-only transport.
1352 Performance patch from Paul Fisher. Bugzilla 1262.
1354 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1357 JH/09 Add $headers_added variable, with content from use of ACL modifier
1358 add_header (but not yet added to the message). Bugzilla 199.
1360 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1361 Pulled from Bugzilla 817 by Wolfgang Breyha.
1363 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1365 (nb: this is the same fix as in Exim 4.80.1)
1367 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1370 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1373 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1375 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1376 advertise SMTP AUTH mechanism to us, instead of a generic
1377 protocol violation error. Also, make Exim more robust to bad
1378 data from the Dovecot auth socket.
1380 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1382 When a queue runner is handling a message, Exim first routes the
1383 recipient addresses, during which it prunes them based on the retry
1384 hints database. After that it attempts to deliver the message to
1385 any remaining recipients. It then updates the hints database using
1388 So if a recipient address works intermittently, it can get repeatedly
1389 deferred at routing time. The retry hints record remains fresh so the
1390 address never reaches the final cutoff time.
1392 This is a fairly common occurrence when a user is bumping up against
1393 their storage quota. Exim had some logic in its local delivery code
1394 to deal with this. However it did not apply to per-recipient defers
1395 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1397 This change adds a proper retry rule check during routing so that the
1398 final cutoff time is checked against the message's age. We only do
1399 this check if there is an address retry record and there is not a
1400 domain retry record; this implies that previous attempts to handle
1401 the address had the retry_use_local_parts option turned on. We use
1402 this as an approximation for the destination being like a local
1403 delivery, as in LMTP.
1405 I suspect this new check makes the old local delivery cutoff check
1406 redundant, but I have not verified this so I left the code in place.
1408 TF/02 Correct gecos expansion when From: is a prefix of the username.
1410 Test 0254 submits a message to Exim with the header
1414 When I ran the test suite under the user fanf2, Exim expanded
1415 the header to contain my full name, whereas it should have added
1416 a Resent-Sender: header. It erroneously treats any prefix of the
1417 username as equal to the username.
1419 This change corrects that bug.
1421 GF/01 DCC debug and logging tidyup
1422 Error conditions log to paniclog rather than rejectlog.
1423 Debug lines prefixed by "DCC: " to remove any ambiguity.
1425 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1427 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1428 Bug spotted by Jeremy Harris; was flawed since initial commit.
1429 Would have resulted in OCSP responses post-SNI triggering an Exim
1430 NULL dereference and crash.
1432 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1434 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1435 Bug detection, analysis and fix by Samuel Thibault.
1436 Bugzilla 1331, Debian bug #698092.
1438 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1440 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1441 Server implementation by Todd Lyons, client by JH.
1442 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1443 config variable "prdr_enable" controls whether the server
1444 advertises the facility. If the client requests PRDR a new
1445 acl_data_smtp_prdr ACL is called once for each recipient, after
1446 the body content is received and before the acl_smtp_data ACL.
1447 The client is controlled by both of: a hosts_try_prdr option
1448 on the smtp transport, and the server advertisement.
1449 Default client logging of deliveries and rejections involving
1450 PRDR are flagged with the string "PRDR".
1452 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1453 fclose(). Diagnosis by Todd Lyons.
1455 PP/17 Update configure.default to handle IPv6 localhost better.
1456 Patch by Alain Williams (plus minor tweaks).
1459 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1460 This is now consistent with GnuTLS, and is now documented: the
1461 previous undocumented portable approach to treating the option as
1462 unset was to force an expansion failure. That still works, and
1463 an empty string is now equivalent.
1465 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1466 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1467 not performing validation itself.
1469 PP/20 Added force_command boolean option to pipe transport.
1470 Patch from Nick Koston, of cPanel Inc.
1472 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1475 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1477 PP/21 Fix eximon continuous updating with timestamped log-files.
1478 Broken in a format-string cleanup in 4.80, missed when I repaired the
1479 other false fix of the same issue.
1480 Report and fix from Heiko Schlichting.
1483 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1484 Report from Prashanth Katuri.
1486 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1487 It's SecureTransport, so affects any MacOS clients which use the
1488 system-integrated TLS libraries, including email clients.
1490 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1491 using a MIME ACL for non-SMTP local injection.
1492 Report and assistance in diagnosis by Warren Baker.
1494 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1496 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1498 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1499 last id that failed so it may be referenced in subsequent ACL's.
1501 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1504 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1505 ldap library initialization, allowing self-signed CA's to be
1506 used. Also properly sets require_cert option later in code by
1507 using NULL (global ldap config) instead of ldap handle (per
1508 session). Bug diagnosis and testing by alxgomz.
1510 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1511 the src/util/ subdirectory.
1513 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1514 renamed to Transport Post Delivery Action by Jeremy Harris, as
1517 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1518 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1519 redis_servers = needs to be configured which will be used by the redis
1520 lookup. Patch from Warren Baker, of The Packet Hub.
1522 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1524 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1525 hostname or reverse DNS when processing a host list. Used suggestions
1526 from multiple comments on this bug.
1528 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1530 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1531 Missed a few lines, added it to make the runtest require no keyboard
1534 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1535 contains upper case chars. Make router use caseful_local_part.
1537 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1538 support when GnuTLS has been built with p11-kit.
1544 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1546 This, or similar/improved, will also be change PP/11 of 4.82.
1552 PP/01 Handle short writes when writing local log-files.
1553 In practice, only affects FreeBSD (8 onwards).
1554 Bugzilla 1053, with thanks to Dmitry Isaikin.
1556 NM/01 Bugzilla 949 - Documentation tweak
1558 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1561 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1563 PP/02 Implemented gsasl authenticator.
1565 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1567 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1568 `pkg-config foo` for cflags/libs.
1570 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1571 with rest of GSASL and with heimdal_gssapi.
1573 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1574 `pkg-config foo` for cflags/libs for the TLS implementation.
1576 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1577 properties get this fed in as external SSF. A number of robustness
1578 and debugging improvements to the cyrus_sasl authenticator.
1580 PP/08 cyrus_sasl server now expands the server_realm option.
1582 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1583 Patch by Jeremy Harris.
1585 PP/10 Added dbmjz lookup type.
1587 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1589 PP/12 MAIL args handles TAB as well as SP, for better interop with
1590 non-compliant senders.
1591 Analysis and variant patch by Todd Lyons.
1593 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1594 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1595 Patch from Dirk Mueller <dmueller@suse.com>
1597 PP/13 tls_peerdn now print-escaped for spool files.
1598 Observed some $tls_peerdn in wild which contained \n, which resulted
1599 in spool file corruption.
1601 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1602 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1603 or write after TLS renegotiation, which otherwise led to messages
1606 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1607 as a tracking header (ie: a signed header comes before the signature).
1608 Patch from Wolfgang Breyha.
1610 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1611 comma-sep list; embedded commas doubled.
1613 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1615 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1617 Report and patch from Dmitry Banschikov.
1619 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1620 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1621 failures. We appear to now support TLS1.1+ with Exim.
1623 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1624 lets Exim select keys and certificates based upon TLS SNI from client.
1625 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1626 before an outbound SMTP session. New log_selector, +tls_sni.
1628 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1629 NULL dereference. Report and patch from Alun Jones.
1631 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1632 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1633 Not seeing resolver debug output on NetBSD, but suspect this is a
1634 resolver implementation change.
1636 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1637 Left warnings. Added "eximon gdb" invocation mode.
1639 PP/21 Defaulting "accept_8bitmime" to true, not false.
1641 PP/22 Added -bw for inetd wait mode support.
1643 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1644 locate the relevant includes and libraries. Made this the default.
1646 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1647 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1649 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1650 This may cause build issues on older platforms.
1652 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1653 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1654 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1655 Added SNI support via GnuTLS too.
1656 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1658 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1660 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1661 Applied second patch from Janne, implementing suggestion to default
1662 multiple-strings-in-record handling to match SPF spec.
1664 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1666 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1667 read-only, out of scope).
1668 Patch from Wolfgang Breyha, report from Stuart Northfield.
1670 PP/29 Fix three issues highlighted by clang analyser static analysis.
1671 Only crash-plausible issue would require the Cambridge-specific
1672 iplookup router and a misconfiguration.
1673 Report from Marcin Mirosław.
1675 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1677 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1678 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1679 As part of this, removing so much warning spew let me fix some minor
1680 real issues in debug logging.
1682 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1683 assignment on my part. Fixed.
1685 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1686 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1687 Janne Snabb (who went above and beyond: thank you).
1689 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1690 string otherwise requires a connection and a bunch more work and it's
1691 relatively easy to get wrong. Should also expose TLS library linkage
1694 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1695 64-bit ${eval} (JH/03).
1697 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1698 GNU libc to support some of the 64-bit stuff, should not lead to
1699 conflicts. Defined before os.h is pulled in, so if a given platform
1700 needs to override this, it can.
1702 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1703 protection layer was required, which is not implemented.
1704 Bugzilla 1254, patch from Wolfgang Breyha.
1706 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1707 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1708 tls_dhparam take prime identifiers. Also unbreak combination of
1709 OpenSSL+DH_params+TLSSNI.
1711 PP/39 Disable SSLv2 by default in OpenSSL support.
1717 PP/01 Solaris build fix for Oracle's LDAP libraries.
1718 Bugzilla 1109, patch from Stephen Usher.
1720 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1722 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1725 TF/02 Fix a couple more cases where we did not log the error message
1726 when unlink() failed. See also change 4.74-TF/03.
1728 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1729 lock up or crash if it happened to be inside a call to libc when it
1730 got a SIGUSR1 from exiwhat.
1732 The SIGUSR1 handler appends the current process status to the process
1733 log which is later printed by exiwhat. It used to use the general
1734 purpose logging code to do this, but several functions it calls are
1735 not safe for signals.
1737 The new output code in the SIGUSR1 handler is specific to the process
1738 log, and simple enough that it's easy to inspect for signal safety.
1739 Removing some special cases also simplifies the general logging code.
1740 Removing the spurious timestamps from the process log simplifies
1743 TF/04 Improved ratelimit ACL condition.
1745 The /noupdate option has been deprecated in favour of /readonly which
1746 has clearer semantics. The /leaky, /strict, and /readonly update modes
1747 are mutually exclusive. The update mode is no longer included in the
1748 database key; it just determines when the database is updated. (This
1749 means that when you upgrade Exim will forget old rate measurements.)
1751 Exim now checks that the per_* options are used with an update mode that
1752 makes sense for the current ACL. For example, when Exim is processing a
1753 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1754 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1755 must specify per_mail/readonly. If you omit the update mode it defaults to
1756 /leaky where that makes sense (as before) or /readonly where required.
1758 The /noupdate option is now undocumented but still supported for
1759 backwards compatibility. It is equivalent to /readonly except that in
1760 ACLs where /readonly is required you may specify /leaky/noupdate or
1761 /strict/noupdate which are treated the same as /readonly.
1763 A useful new feature is the /count= option. This is a generalization
1764 of the per_byte option, so that you can measure the throughput of other
1765 aggregate values. For example, the per_byte option is now equivalent
1766 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1768 The per_rcpt option has been generalized using the /count= mechanism
1769 (though it's more complicated than the per_byte equivalence). When it is
1770 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1771 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1772 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1773 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1774 Note that using per_rcpt with a non-readonly update mode in more than
1775 one ACL will cause the recipients to be double-counted. (The per_mail
1776 and per_byte options don't have this problem.)
1778 The handling of very low rates has changed slightly. If the computed rate
1779 is less than the event's count (usually one) then this event is the first
1780 after a long gap. In this case the rate is set to the same as this event's
1781 count, so that the first message of a spam run is counted properly.
1783 The major new feature is a mechanism for counting the rate of unique
1784 events. The new per_addr option counts the number of different
1785 recipients that someone has sent messages to in the last time period. It
1786 behaves like per_rcpt if all the recipient addresses are different, but
1787 duplicate recipient addresses do not increase the measured rate. Like
1788 the /count= option this is a general mechanism, so the per_addr option
1789 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1790 example, measure the rate that a client uses different sender addresses
1791 with the options per_mail/unique=$sender_address. There are further
1792 details in the main documentation.
1794 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1796 TF/06 Removed a few PCRE remnants.
1798 TF/07 Automatically extract Exim's version number from tags in the git
1799 repository when doing development or release builds.
1801 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1802 Bugzilla 879. Patch from Paul Fisher.
1804 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1805 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1808 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1810 PP/05 Variable $av_failed, true if the AV scanner deferred.
1811 Bugzilla 1078. Patch from John Horne.
1813 PP/06 Stop make process more reliably on build failure.
1814 Bugzilla 1087. Patch from Heiko Schlittermann.
1816 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1817 Bugzilla 1089. Patch from Heiko Schlittermann.
1819 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1820 Bugzilla 1131. Patch from Holger Weiß.
1822 PP/09 Handle IPv6 addresses with SPF.
1823 Bugzilla 860. Patch from Wolfgang Breyha.
1825 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1827 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1830 PP/11 match_* no longer expand right-hand-side by default.
1831 New compile-time build option, EXPAND_LISTMATCH_RHS.
1832 New expansion conditions, "inlist", "inlisti".
1834 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1836 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1838 PP/14 fix log_write() format string regression from TF/03.
1839 Bugzilla 1152. Patch from Dmitry Isaikin.
1845 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1847 PP/02 Harmonised TLS library version reporting; only show if debugging.
1848 Layout now matches that introduced for other libraries in 4.74 PP/03.
1850 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1852 PP/04 New "dns_use_edns0" global option.
1854 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1857 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1858 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1860 TK/01 Updated PolarSSL code to 0.14.2.
1861 Bugzilla 1097. Patch from Andreas Metzler.
1863 PP/07 Catch divide-by-zero in ${eval:...}.
1864 Fixes bugzilla 1102.
1866 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1869 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1870 format-string attack -- SECURITY: remote arbitrary code execution.
1872 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1873 time unintentionally subject to list matching rules, letting the header
1874 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1875 arbitrary string expansion). This allowed for information disclosure.
1877 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1878 INT_MIN/-1 -- value coerced to INT_MAX.
1884 NM/01 Workaround for PCRE version dependency in version reporting
1887 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1888 This fixes portability to compilers other than gcc, notably
1889 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1891 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1892 makefiles for portability to HP-UX and POSIX correctness.
1894 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1895 Also via indented variable definition in the Makefile.
1896 (Debugging by Oliver Heesakkers).
1898 PP/02 Restore caching of spamd results with expanded spamd_address.
1899 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1901 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1902 Improves build reliability. Fix from: Frank Elsner
1904 NM/02 Fix wide character breakage in the rfc2047 coding
1905 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1907 NM/03 Allow underscore in dnslist lookups
1908 Fixes bug 1026. Patch from Graeme Fowler
1910 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1911 Code patches from Adam Ciarcinski of NetBSD.
1913 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1916 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1917 is logged, to avoid truncation. Patch from John Horne.
1919 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1920 Patch from Jakob Hirsch.
1922 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1923 SQL string expansion failure details.
1924 Patch from Andrey Oktyabrski.
1926 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1927 Patch from Simon Arlott.
1929 PP/09 New lookups functionality failed to compile on old gcc which rejects
1930 extern declarations in function scope.
1931 Patch from Oliver Fleischmann
1933 PP/10 Use sig_atomic_t for flags set from signal handlers.
1934 Check getgroups() return and improve debugging.
1935 Fixed developed for diagnosis in bug 927 (which turned out to be
1938 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1939 Patch from Mark Zealey.
1941 PP/12 Bugzilla 1056: Improved spamd server selection.
1942 Patch from Mark Zealey.
1944 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1945 Based on patch from Heiko Schlittermann.
1947 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1948 Patch from Uwe Doering, sign-off by Michael Haardt.
1950 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1951 variable declaration deep within a block. Bug and patch from
1954 PP/15 lookups-Makefile IRIX compatibility coercion.
1956 PP/16 Make DISABLE_DKIM build knob functional.
1958 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1959 Patch by Simon Arlott
1961 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1962 variable argument macros. Our copy now differs from upstream.
1968 TF/01 Failure to get a lock on a hints database can have serious
1969 consequences so log it to the panic log.
1971 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1972 controlled using the smtp_confirmation log selector.
1974 TF/03 Include the error message when we fail to unlink a spool file.
1976 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1977 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1978 for maintaining out-of-tree patches for some time.
1980 PP/01 Bugzilla 139: Documentation and portability issues.
1981 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1982 Handle per-OS dynamic-module compilation flags.
1984 PP/02 Let /dev/null have normal permissions.
1985 The 4.73 fixes were a little too stringent and complained about the
1986 permissions on /dev/null. Exempt it from some checks.
1987 Reported by Andreas M. Kirchwitz.
1989 PP/03 Report version information for many libraries, including
1990 Exim version information for dynamically loaded libraries. Created
1991 version.h, now support a version extension string for distributors
1992 who patch heavily. Dynamic module ABI change.
1994 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1995 privilege escalation vulnerability whereby the Exim run-time user
1996 can cause root to append content of the attacker's choosing to
1999 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2002 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2003 If dropping privileges for untrusted macros, we disabled normal logging
2004 on the basis that it would fail; for the Exim run-time user, this is not
2005 the case, and it resulted in successful deliveries going unlogged.
2006 Fixed. Reported by Andreas Metzler.
2012 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2013 only prepend for the Resent-* case. Fixes regression introduced in
2014 Exim 4.70 by NM/22 for Bugzilla 607.
2016 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2017 increasing numbers of administrators be bitten by this.
2019 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2021 PP/03 Bugzilla 994: added openssl_options main configuration option.
2023 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2025 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2027 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2029 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2030 without a peer certificate, leading to a segfault because of an
2031 assumption that peers always have certificates. Be a little more
2032 paranoid. Problem reported by Martin Tscholak.
2034 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2035 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2036 NB: ClamAV planning to remove STREAM in "middle of 2010".
2037 CL also introduces -bmalware, various -d+acl logging additions and
2038 more caution in buffer sizes.
2040 PP/09 Implemented reverse_ip expansion operator.
2042 PP/10 Bugzilla 937: provide a "debug" ACL control.
2044 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2046 PP/12 Bugzilla 973: Implement --version.
2048 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2050 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2052 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2054 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2055 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2056 ignore trailing whitespace.
2058 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2060 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2063 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2064 Notification from Dr Andrew Aitchison.
2066 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2067 ExtendedDetectionInfo response format.
2068 Notification from John Horne.
2070 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2073 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2074 XSL and documented dependency on system catalogs, with examples of how
2077 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2080 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2081 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2082 configuration file which is writeable by the Exim user or group.
2084 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2085 of configuration files to cover files specified with the -C option if
2086 they are going to be used with root privileges, not just the default
2089 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2090 option (effectively making it always true).
2092 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2093 files to be used while preserving root privileges.
2095 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2096 that rogue child processes cannot use them.
2098 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2099 run-time user, instead of root.
2101 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2102 Exim run-time user without dropping privileges.
2104 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2105 result string, instead of calling string_vformat() twice with the same
2108 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2109 for other users. Others should always drop root privileges if they use
2110 -C on the command line, even for a whitelisted configure file.
2112 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2114 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2120 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2121 $data_path, and $header_path variables; fixed documentation bugs and
2124 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2125 exipick to access non-standard spools, including the "frozen" queue
2128 NM/01 Bugzilla 965: Support mysql stored procedures.
2129 Patch from Alain Williams
2131 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2133 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2134 Patch from Andreas Metzler
2136 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2137 Patch from Kirill Miazine
2139 NM/05 Bugzilla 671: Added umask to procmail example.
2141 JJ/03 installed exipick 20100323.0, fixing doc bug
2143 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2144 directory. Notification and patch from Dan Rosenberg.
2146 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2148 TK/02 Improve log output when DKIM signing operation fails.
2150 MH/01 Treat the transport option dkim_domain as a colon separated
2151 list, not as a single string, and sign the message with each element,
2152 omitting multiple occurences of the same signer.
2154 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2155 Bugzilla 985, 986. Patch by Simon Arlott
2157 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2158 Patch by Simon Arlott
2160 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2161 MBX locking. Notification from Dan Rosenberg.
2167 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2169 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2171 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2173 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2175 NM/04 Bugzilla 915: Flag broken perl installation during build.
2181 TK/01 Added patch by Johannes Berg that expands the main option
2182 "spamd_address" if it starts with a dollar sign.
2184 TK/02 Write list of recipients to X-Envelope-Sender header when building
2185 the mbox-format spool file for content scanning (suggested by Jakob
2188 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2189 (http://www.dcc-servers.net/) support via dccifd. Activated by
2190 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2192 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2193 by Mark Daniel Reidel <mr@df.eu>.
2195 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2196 When building exim an external PCRE library is now needed -
2197 PCRE is a system library on the majority of modern systems.
2198 See entry on PCRE_LIBS in EDITME file.
2200 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2201 conversation. Added nologin parameter to request.
2202 Patch contributed by Kirill Miazine.
2204 TF/01 Do not log submission mode rewrites if they do not change the address.
2206 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2208 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2209 log files in place. Contributed by Roberto Lima.
2211 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2213 TF/03 Bugzilla 615: When checking the local_parts router precondition
2214 after a local_part_suffix or local_part_prefix option, Exim now
2215 does not use the address's named list lookup cache, since this
2216 contains cached lookups for the whole local part.
2218 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2219 Robert Millan. Documentation is in experimental-spec.txt.
2221 TF/04 Bugzilla 668: Fix parallel build (make -j).
2223 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2225 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2226 Patch provided by Jan Srzednicki.
2228 TF/05 Leading white space used to be stripped from $spam_report which
2229 wrecked the formatting. Now it is preserved.
2231 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2232 that they are available at delivery time.
2234 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2236 TF/08 TLS error reporting now respects the incoming_interface and
2237 incoming_port log selectors.
2239 TF/09 Produce a more useful error message if an SMTP transport's hosts
2240 setting expands to an empty string.
2242 NM/06 Bugzilla 744: EXPN did not work under TLS.
2243 Patch provided by Phil Pennock.
2245 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2246 Patch provided by Richard Godbee.
2248 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2249 acl_smtp_notquit, added index entry.
2251 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2252 Patch provided by Eugene Bujak.
2254 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2255 accept(). Patch provided by Maxim Dounin.
2257 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2258 Patch provided by Phil Pennock.
2260 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2262 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2263 Patch provided by Brad "anomie" Jorsch.
2265 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2266 Patch provided by Dean Brooks.
2268 TK/05 Add native DKIM support (does not depend on external libraries).
2270 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2271 Patch provided by Graeme Fowler.
2273 NM/16 Bugzilla 851: Documentation example syntax fix.
2275 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2277 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2280 NM/19 Bugzilla 745: TLS version reporting.
2281 Patch provided by Phil Pennock.
2283 NM/20 Bugzilla 167: bool: condition support.
2284 Patch provided by Phil Pennock.
2286 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2287 clients. Patch provided by Phil Pennock.
2289 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2290 Patch provided by Brad "anomie" Jorsch.
2292 NM/23 Bugzilla 687: Fix misparses in eximstats.
2293 Patch provided by Heiko Schlittermann.
2295 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2296 Patch provided by Heiko Schlittermann.
2298 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2299 plus update to original patch.
2301 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2303 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2304 Patch provided by David Brownlee.
2306 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2308 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2310 NM/30 Bugzilla 888: TLS documentation bugfixes.
2312 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2314 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2315 Unlike the original bugzilla I have changed all shell scripts in src tree.
2317 NM/33 Bugzilla 898: Transport filter timeout fix.
2318 Patch by Todd Rinaldo.
2320 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2321 Patch by Serge Demonchaux.
2323 NM/35 Bugzilla 39: Base64 decode bug fixes.
2324 Patch by Jakob Hirsch.
2326 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2328 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2330 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2332 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2338 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2339 ALT-N's libdkim that I have put here:
2340 http://duncanthrax.net/exim-experimental/
2342 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2343 were called 'true' and it seems that C99 defines that as a reserved
2344 keyword to be used with 'bool' variable types. That means you could
2345 not include C99-style headers which use bools without triggering
2346 build errors in sieve.c.
2348 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2349 as mailq or other aliases. Changed the --help handling significantly
2350 to do whats expected. exim_usage() emits usage/help information.
2352 SC/01 Added the -bylocaldomain option to eximstats.
2354 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2356 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2358 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2364 PH/01 Another patch from the Sieve maintainer.
2366 PH/02 When an IPv6 address is converted to a string for single-key lookup
2367 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2368 dots are used instead of colons so that keys in lsearch files need not
2369 contain colons. This was done some time before quoting was made available
2370 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2371 (notated using the quote facility) so as to distinguish them from IPv4
2372 keys. This meant that lookups for IP addresses in host lists did not work
2373 for iplsearch lookups.
2375 This has been fixed by arranging for IPv6 addresses to be expressed with
2376 colons if the lookup type is iplsearch. This is not incompatible, because
2377 previously such lookups could never work.
2379 The situation is now rather anomalous, since one *can* have colons in
2380 ordinary lsearch keys. However, making the change in all cases is
2381 incompatible and would probably break a number of configurations.
2383 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2386 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2387 conversion specification without a maximum field width, thereby enabling
2388 a rogue spamd server to cause a buffer overflow. While nobody in their
2389 right mind would setup Exim to query an untrusted spamd server, an
2390 attacker that gains access to a server running spamd could potentially
2391 exploit this vulnerability to run arbitrary code as the Exim user.
2393 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2394 $primary_hostname instead of what libspf2 thinks the hosts name is.
2396 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2397 a directory entry by the name of the lookup key. Previously, if a
2398 symlink pointed to a non-existing file or a file in a directory that
2399 Exim lacked permissions to read, a lookup for a key matching that
2400 symlink would fail. Now it is enough that a matching directory entry
2401 exists, symlink or not. (Bugzilla 503.)
2403 PH/03 The body_linecount and body_zerocount variables are now exported in the
2406 PH/04 Added the $dnslist_matched variable.
2408 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2409 This means they are set thereafter only if the connection becomes
2412 PH/06 Added the client_condition to authenticators so that some can be skipped
2413 by clients under certain conditions.
2415 PH/07 The error message for a badly-placed control=no_multiline_responses left
2416 "_responses" off the end of the name.
2418 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2420 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2421 (without spaces) instead of just copying the configuration text.
2423 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2425 PH/11 Added $max_received_linelength.
2427 PH/12 Added +ignore_defer and +include_defer to host lists.
2429 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2430 way in which PCRE > 7.0 is built.
2432 PH/14 Implemented queue_only_load_latch.
2434 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2435 MAIL command. The effect was to mangle the value on 64-bit systems.
2437 PH/16 Another patch from the Sieve maintainer.
2439 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2441 PH/18 If a system quota error occurred while trying to create the file for
2442 a maildir delivery, the message "Mailbox is full" was not appended to the
2443 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2444 only to a quota excession during the actual writing of the file.
2446 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2447 characters?) which causes problems in log lines. The DN values are now
2448 passed through string_printing() before being added to log lines.
2450 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2451 and InterBase are left for another time.)
2453 PH/21 Added message_body_newlines option.
2455 PH/22 Guard against possible overflow in moan_check_errorcopy().
2457 PH/23 POSIX allows open() to be a macro; guard against that.
2459 PH/24 If the recipient of an error message contained an @ in the local part
2460 (suitably quoted, of course), incorrect values were put in $domain and
2461 $local_part during the evaluation of errors_copy.
2467 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2468 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2471 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2472 issue a MAIL command.
2474 PH/02 In an ACL statement such as
2476 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2478 if a client was not listed at all, or was listed with a value other than
2479 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2480 the condition was not true (as it should be), so access was not denied.
2481 The bug was that the ! inversion was incorrectly passed on to the second
2482 item. This has been fixed.
2484 PH/03 Added additional dnslists conditions == and =& which are different from
2485 = and & when the dns lookup returns more than one IP address.
2487 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2488 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2490 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2491 FSYNC, which compiles an option called disable_fsync that allows for
2492 bypassing fsync(). The documentation is heavily laced with warnings.
2494 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2496 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2497 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2498 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2499 including adding "make clean"; (3) Added -fPIC when compiling the test
2500 dynamically loaded module, to get rid of a warning.
2502 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2503 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2504 = 0s. The bug is otherwise harmless.
2506 PH/07 There was a bug in the dovecot authenticator such that the value of
2507 $auth1 could be overwritten, and so not correctly preserved, after a
2508 successful authentication. This usually meant that the value preserved by
2509 the server_setid option was incorrect.
2511 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2513 PH/09 Installed PCRE release 7.0.
2515 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2516 run for batched SMTP input. It is now run at the start of every message
2517 in the batch. While fixing this I discovered that the process information
2518 (output by running exiwhat) was not always getting set for -bs and -bS
2519 input. This is fixed, and it now also says "batched" for BSMTP.
2521 PH/11 Added control=no_pipelining.
2523 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2524 patch, slightly modified), and move the expansion of helo_data till after
2525 the connection is made in the smtp transport (so it can use these
2528 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2530 PH/14 Added log_selector = +pid.
2532 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2534 PH/16 Add ${if forany and ${if forall.
2536 PH/17 Added dsn_from option to vary the From: line in DSNs.
2538 PH/18 Flush SMTP output before performing a callout, unless control =
2539 no_callout_flush is set.
2541 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2542 was true (the default) a successful delivery failed to delete the retry
2543 item, thus causing premature timeout of the address. The bug is now
2546 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2548 PH/21 Long custom messages for fakedefer and fakereject are now split up
2549 into multiline responses in the same way that messages for "deny" and
2550 other ACL rejections are.
2552 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2553 with slight modification.
2555 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2556 draft, changing the syntax and factoring some duplicate code".
2558 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2559 for deliveries of the second and subsequent messages over the same SMTP
2562 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2563 ${reduce, with only minor "tidies".
2565 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2567 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2568 expansion side effects.
2570 PH/27 When a message times out after an over-quota error from an Exim-imposed
2571 quota, the bounce message says "mailbox is full". This message was not
2572 being given when it was a system quota that was exceeded. It now should
2575 MH/03 Made $recipients available in local_scan(). local_scan() already has
2576 better access to the recipient list through recipients_list[], but
2577 $recipients can be useful in postmaster-provided expansion strings.
2579 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2580 in the case of a MAIL command with additional options following the
2581 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2582 were accidentally chopped off.
2584 PH/29 SMTP synchronization checks are implemented when a command is read -
2585 there is a check that no more input is waiting when there shouldn't be
2586 any. However, for some commands, a delay in an ACL can mean that it is
2587 some time before the response is written. In this time, more input might
2588 arrive, invalidly. So now there are extra checks after an ACL has run for
2589 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2590 pipelining has not been advertised.
2592 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2594 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2595 returned in the SMTP response when smtp_return_error_details was false.
2596 This has been fixed.
2598 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2599 instead of the C I/O that was originally supplied, because problems were
2600 reported on Solaris.
2602 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2603 Exim which did not show up earlier: it was assuming that a call to
2604 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2605 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2606 was a macro that became an assignment, so it seemed to work. This has
2607 changed to a proper function call with a void return, hence the compile
2608 error. Exim's code has been fixed.
2610 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2613 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2615 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2617 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2618 because it turns out to be too big or there is a timeout) there is no
2619 "Completed" line in the log. When some messages of this type were
2620 selected by exigrep, they were listed as "not completed". Others were
2621 picked up by some special patterns. I have improved the selection
2622 criteria to be more general.
2624 PH/38 The host_find_failed option in the manualroute router can now be set
2625 to "ignore", to completely ignore a host whose IP address cannot be
2626 found. If all hosts are ignored, the behaviour is controlled by the new
2627 host_all_ignored option.
2629 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2630 homing or because of multiple MX records with /mx) generated more than
2631 one IP address, and the following item turned out to be the local host,
2632 all the secondary addresses of the first item were incorrectly removed
2633 from the list, along with the local host and any following hosts (which
2634 is what is supposed to happen).
2636 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2637 whoever called Exim into the -H file. In the case of the daemon it was
2638 behaving confusingly. When first started, it used values for whoever
2639 started the daemon, but after a SIGHUP it used the Exim user (because it
2640 calls itself on a restart). I have changed the code so that it now always
2643 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2644 message are rejected with the same error (e.g. no authentication or bad
2645 sender address), and a DATA command is nevertheless sent (as can happen
2646 with PIPELINING or a stupid MUA), the error message that was given to the
2647 RCPT commands is included in the rejection of the DATA command. This is
2648 intended to be helpful for MUAs that show only the final error to their
2651 PH/42 Another patch from the Sieve maintainer.
2653 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2654 Eximstats - Fixed some broken HTML links and added missing column headers
2656 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2657 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2659 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2661 PH/43 Yet another patch from the Sieve maintainer.
2663 PH/44 I found a way to check for a TCP/IP connection going away before sending
2664 the response to the final '.' that terminates a message, but only in the
2665 case where the client has not sent further data following the '.'
2666 (unfortunately, this is allowed). However, in many cases there won't be
2667 any further data because there won't be any more messages to send. A call
2668 to select() can be used: if it shows that the input is "ready", there is
2669 either input waiting, or the socket has been closed. An attempt to read
2670 the next input character can distinguish the two cases. Previously, Exim
2671 would have sent an OK response which the client would never have see.
2672 This could lead to message repetition. This fix should cure that, at
2673 least in a lot of common cases.
2675 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2676 advertised in response to EHLO.
2682 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2683 fixed by 4.65/MH/01 (is this a record?) are fixed:
2685 (i) An empty string was always treated as zero by the numeric comparison
2686 operators. This behaviour has been restored.
2688 (ii) It is documented that the numeric comparison operators always treat
2689 their arguments as decimal numbers. This was broken in that numbers
2690 starting with 0 were being interpreted as octal.
2692 While fixing these problems I realized that there was another issue that
2693 hadn't been noticed. Values of message_size_limit (both the global option
2694 and the transport option) were treated as octal if they started with 0.
2695 The documentation was vague. These values are now always treated as
2696 decimal, and I will make that clear in the documentation.
2702 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2703 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2706 MH/01 Don't check that the operands of numeric comparison operators are
2707 integers when their expansion is in "skipping" mode (fixes bug
2708 introduced by 4.64-PH/07).
2710 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2711 child addresses, Exim now panics and dies. Previously, because the count
2712 is held in a short int, deliveries were likely to be lost. As such a
2713 large number of recipients for a single message is ridiculous
2714 (performance will be very, very poor), I have chosen to impose a limit
2715 rather than extend the field.
2721 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2722 leftover -K file (the existence of which was triggered by #402).
2723 While we were at it, introduced process PID as part of the -K
2724 filename. This should rule out race conditions when creating
2727 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2728 processing considerably. Previous code took too long for large mails,
2729 triggering a timeout which in turn triggers #401.
2731 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2732 in the DK code in transports.c. sendfile() is not really portable,
2733 hence the _LINUX specificness.
2735 TF/01 In the add_headers option to the mail command in an Exim filter,
2736 there was a bug that Exim would claim a syntax error in any
2737 header after the first one which had an odd number of characters
2740 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2741 callout verification, Exim cached a "reject" for the entire domain. This
2742 is correct for most verifications, but it is not correct for a recipient
2743 verification with use_sender or use_postmaster set, because in that case
2744 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2745 case of MAIL FROM:<> rejection from other early rejections (e.g.
2746 rejection of HELO). When verifying a recipient using a non-null MAIL
2747 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2748 Whatever the result of the callout, the value of the domain cache is
2749 left unchanged (for any other kind of callout, getting as far as trying
2750 RCPT means that the domain itself is ok).
2752 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2755 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2756 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2757 session, the daemon ignores ECONNECT errors and logs others; it now
2758 ignores EPIPE as well.
2760 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2761 (quoted-printable decoding).
2763 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2764 later the small subsequent patch to fix an introduced bug.
2766 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2768 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2770 PH/08 An error is now given if message_size_limit is specified negative.
2772 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2773 to be given (somewhat) arbitrary names.
2775 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2778 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2779 miscellaneous code fixes
2781 PH/10 Added the log_reject_target ACL modifier to specify where to log
2784 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2785 hostname. This is wrong, because it relates to the incoming message (and
2786 probably the interface on which it is arriving) and not to the outgoing
2787 callout (which could be using a different interface). This has been
2788 changed to use the value of the helo_data option from the smtp transport
2789 instead - this is what is used when a message is actually being sent. If
2790 there is no remote transport (possible with a router that sets up host
2791 addresses), $smtp_active_hostname is used.
2793 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2794 tweaks were necessary in order to get it to work (see also 21 below):
2795 (a) The code assumed that strncpy() returns a negative number on buffer
2796 overflow, which isn't the case. Replaced with Exim's string_format()
2798 (b) There were several signed/unsigned issues. I just did the minimum
2799 hacking in of casts. There is scope for a larger refactoring.
2800 (c) The code used strcasecmp() which is not a standard C function.
2801 Replaced with Exim's strcmpic() function.
2802 (d) The code set only $1; it now sets $auth1 as well.
2803 (e) A simple test gave the error "authentication client didn't specify
2804 service in request". It would seem that Dovecot has changed its
2805 interface. Fortunately there's a specification; I followed it and
2806 changed what the client sends and it appears to be working now.
2808 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2811 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2812 address A is aliased to B and C, where B exists and C does not. Without
2813 -v the output is "A verified" because verification stops after a
2814 successful redirection if more than one address is generated. However,
2815 with -v the child addresses are also verified. Exim was outputting "A
2816 failed to verify" and then showing the successful verification for C,
2817 with its parentage. It now outputs "B failed to verify", showing B's
2818 parentage before showing the successful verification of C.
2820 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2821 look up a TXT record in a specific list after matching in a combined
2824 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2825 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2826 they consult the DNS. I had assumed they would set it the way they
2827 wanted; and indeed my experiments on Linux seem to show that in some
2828 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2829 To be on the safe side, however, I have now made the interface to
2830 host_find_byname() similar to host_find_bydns(), with an argument
2831 containing the DNS resolver options. The host_find_byname() function now
2832 sets these options at its start, just as host_find_bydns() does. The smtp
2833 transport options dns_qualify_single and dns_search_parents are passed to
2834 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2835 of host_find_byname() use the default settings of RES_DEFNAMES
2836 (qualify_single) but not RES_DNSRCH (search_parents).
2838 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2839 spool_read_header() do less string testing, by means of a preliminary
2840 switch on the second character of optional "-foo" lines. (This is
2841 overdue, caused by the large number of possibilities that now exist.
2842 Originally there were few.) While I was there, I also converted the
2843 str(n)cmp tests so they don't re-test the leading "-" and the first
2844 character, in the hope this might squeeze out yet more improvement.
2846 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2847 flag allowing group syntax was set by the header_syntax check but not
2848 turned off, possible causing trouble later; (2) The flag was not being
2849 set at all for the header_verify test, causing "group"-style headers to
2850 be rejected. I have now set it in this case, and also caused header_
2851 verify to ignore an empty address taken from a group. While doing this, I
2852 came across some other cases where the code for allowing group syntax
2853 while scanning a header line wasn't quite right (mostly, not resetting
2854 the flag correctly in the right place). These bugs could have caused
2855 trouble for malformed header lines. I hope it is now all correct.
2857 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2858 with the "reply" argument non-NULL. The code, however (which originally
2859 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2860 but it didn't always do it. This confused somebody who was copying the
2861 code for some other use. I have removed all the tests.
2863 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2864 feature that was used to support insecure browsers during the U.S. crypto
2865 embargo. It requires special client support, and Exim is probably the
2866 only MTA that supported it -- and would never use it because real RSA is
2867 always available. This code has been removed, because it had the bad
2868 effect of slowing Exim down by computing (never used) parameters for the
2869 RSA_EXPORT functionality.
2871 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2872 authenticator to fail if there's a tab character in the incoming data
2873 (there should never be unless someone is messing about, as it's supposed
2874 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2875 if the connection is using TLS or if the remote IP is the same as the
2876 local IP, and the "valid-client-cert option" if a client certificate has
2879 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2880 authenticators. This can be used for authorization after authentication
2881 succeeds. (In the case of plaintext, it servers for both authentication
2884 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2885 if any retry times were supplied.
2887 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2888 connection, where there is no client IP address to check. In this
2889 situation, the verify now always succeeds.
2891 PH/25 Applied John Jetmore's -Mset patch.
2893 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2895 PH/27 In a string expansion for a processed (not raw) header when multiple
2896 headers of the same name were present, leading whitespace was being
2897 removed from all of them, but trailing whitespace was being removed only
2898 from the last one. Now trailing whitespace is removed from each header
2899 before concatenation. Completely empty headers in a concatenation (as
2900 before) are ignored.
2902 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2903 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2905 PH/29 [Removed. This was a change that I later backed out, and forgot to
2906 correct the ChangeLog entry (that I had efficiently created) before
2907 committing the later change.]
2909 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2910 address errors (4xx response to RCPT) over the same connection as other
2911 messages routed to the same hosts. Such deliveries are always "forced",
2912 so retry times are not inspected. This resulted in far too many retries
2913 for the affected addresses. The effect occurred only when there were more
2914 hosts than the hosts_max_try setting in the smtp transport when it had
2915 the 4xx errors. Those hosts that it had tried were not added to the list
2916 of hosts for which the message was waiting, so if all were tried, there
2917 was no problem. Two fixes have been applied:
2919 (i) If there are any address or message errors in an SMTP delivery, none
2920 of the hosts (tried or untried) are now added to the list of hosts
2921 for which the message is waiting, so the message should not be a
2922 candidate for sending over the same connection that was used for a
2923 successful delivery of some other message. This seems entirely
2924 reasonable: after all the message is NOT "waiting for some host".
2925 This is so "obvious" that I'm not sure why it wasn't done
2926 previously. Hope I haven't missed anything, but it can't do any
2927 harm, as the worst effect is to miss an optimization.
2929 (ii) If, despite (i), such a delivery is accidentally attempted, the
2930 routing retry time is respected, so at least it doesn't keep
2931 hammering the server.
2933 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2934 in ${readsocket because some servers need this prod.
2936 PH/32 Added some extra debug output when updating a wait-xxx database.
2938 PH/33 The hint "could be header name not terminated by colon", which has been
2939 given for certain expansion errors for a long time, was not being given
2940 for the ${if def:h_colon_omitted{... case.
2942 PH/34 The spec says: "With one important exception, whenever a domain list is
2943 being scanned, $domain contains the subject domain." There was at least
2944 one case where this was not true.
2946 PH/35 The error "getsockname() failed: connection reset by peer" was being
2947 written to the panic log as well as the main log, but it isn't really
2948 panic-worthy as it just means the connection died rather early on. I have
2949 removed the panic log writing for the ECONNRESET error when getsockname()
2952 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2953 runs only) independently of the message's sender address. This meant
2954 that, if the 4xx error was in fact related to the sender, a different
2955 message to the same recipient with a different sender could confuse
2956 things. In particular, this can happen when sending to a greylisting
2957 server, but other circumstances could also provoke similar problems.
2958 I have changed the default so that the retry time for these errors is now
2959 based a combination of the sender and recipient addresses. This change
2960 can be overridden by setting address_retry_include_sender=false in the
2963 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2964 remote server are returned as part of bounce messages. This was not
2965 happening for LMTP over a pipe (the lmtp transport), but now it is the
2966 same for both kinds of LMTP.
2968 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2969 in header lines that were in fact CNAMEs. This is no longer the case.
2971 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2972 and queue runs started by the daemon processed all messages. This has
2973 been fixed so that -R and -S can now usefully be given with -q<time>.
2975 PH/40 Import PCRE release 6.7 (fixes some bugs).
2977 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2979 PH/42 Give an error if -q is specified more than once.
2981 PH/43 Renamed the variables $interface_address and $interface_port as
2982 $received_ip_address and $received_port, to make it clear that these
2983 values apply to message reception, and not to the outgoing interface when
2984 a message is delivered. (The old names remain recognized, of course.)
2986 PH/44 There was no timeout on the connect() call when using a Unix domain
2987 socket in the ${readsocket expansion. There now is.
2989 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2990 be meaningful with "accept".
2992 SC/01 Eximstats V1.43
2993 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2995 SC/02 Eximstats V1.44
2996 Use a glob alias rather than an array ref in the generated
2997 parser. This improves both readability and performance.
2999 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3000 Collect SpamAssassin and rejection statistics.
3001 Don't display local sender or destination tables unless
3002 there is data to show.
3003 Added average volumes into the top table text output.
3005 SC/04 Eximstats V1.46
3006 Collect data on the number of addresses (recipients)
3007 as well as the number of messages.
3009 SC/05 Eximstats V1.47
3010 Added 'Message too big' to the list of mail rejection
3011 reasons (thanks to Marco Gaiarin).
3013 SC/06 Eximstats V1.48
3014 Mainlog lines which have GMT offsets and are too short to
3015 have a flag are now skipped.
3017 SC/07 Eximstats V1.49 (Alain Williams)
3018 Added the -emptyok flag.
3020 SC/08 Eximstats V1.50
3021 Fixes for obtaining the IP address from reject messages.
3023 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3024 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3025 whitespace changes from 4.64-PH/27
3027 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3030 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3031 are found, allow negative numbers in numeric criteria)
3033 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3035 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3036 to match changes made in 4.64-PH/43
3038 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3040 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3041 "domain" is actually an IP address, and force a failure. This locks out
3042 those revolvers/nameservers that support "A-for-A" lookups, in
3043 contravention of the specifications.
3045 PH/48 When a host name was looked up from an IP address, and the subsequent
3046 forward lookup of the name timed out, the host name was left in
3047 $sender_host_name, contrary to the specification.
3049 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3050 restricted to single-key lookups, Exim was not diagnosing an error if
3051 * or *@ was used with a query-style lookup.
3053 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3055 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3056 long ago, but noone interested enough thought of it. Let's just say that
3057 the "1.1" means that there are some new functions that weren't there at
3058 some point in the past.
3060 PH/51 Error processing for expansion failure of helo_data from an smtp
3061 transport during callout processing was broken.
3063 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3064 tested/used via the -bh/-bhc/-bs options.
3066 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3067 bug, fixed in subsequent PCRE releases).
3069 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3070 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3072 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3078 SC/01 Use a glob alias rather than an array ref in eximstats generated
3079 parser. This improves both readability and performance.
3081 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3082 Don't display local sender or destination tables in eximstats unless
3083 there is data to show.
3084 Added average volumes into the eximstats top table text output.
3086 SC/03 Collect data on the number of addresses (recipients) as well
3087 as the number of messages in eximstats.
3089 TF/01 Correct an error in the documentation for the redirect router. Exim
3090 does (usually) call initgroups() when daemonizing.
3092 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3093 with consistent privilege compared to when running as a daemon.
3095 TF/03 Note in the spec that $authenticated_id is not set for local
3096 submissions from trusted users.
3098 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3099 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3101 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3102 by adding some example configuration directives to the default
3103 configuration file. A little bit of work is required to uncomment the
3104 directives and define how usernames and passwords are checked, but
3105 there is now a framework to start from.
3107 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3108 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3109 without this. I don't know how relevant this is to other LDAP libraries.
3111 PH/02 Add the verb name to the "unknown ACL verb" error.
3113 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3115 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3117 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3118 directory not expanded when it should be if an expanded home directory
3119 was set for the address (which is overridden by the transport).
3121 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3124 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3125 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3126 because it is too late at that time, and has no effect.
3128 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3129 security issue with \' (bugzilla #107). I could not use the
3130 PQescapeStringConn() function, because it needs a PGconn value as one of
3133 PH/08 When testing addresses using -bt, indicate those final addresses that
3134 are duplicates that would not cause an additional delivery. At least one
3135 person was confused, thinking that -bt output corresponded to deliveries.
3136 (Suppressing duplicates isn't a good idea as you lose the information
3137 about possibly different redirections that led to the duplicates.)
3139 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3140 systems where poll() doesn't work, in particular OS X.
3142 PH/10 Added more information to debugging output for retry time not reached.
3144 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3145 operations in malware.c.
3147 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3150 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3151 syslog_duplication set false, log lines that would normally be written
3152 both the the main log and to the reject log were not written to syslog at
3155 PH/14 In the default configuration, change the use of "message" in ACL warn
3156 statements to "add_header".
3158 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3159 not followed by a command (e.g. "seen endif").
3161 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3162 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3165 PH/17 Added extra conditions to the default value of delay_warning_condition
3169 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3170 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3171 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3174 The Auto-Submitted: and various List- headers are standardised, whereas I
3175 don't think Precedence: ever was.
3177 PH/18 Refactored debugging code in route_finduser() to show more information,
3178 in particular, the error code if getpwnam() issues one.
3180 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3181 This is apparently needed in addition to the PH/07 change above to avoid
3182 any possible encoding problems.
3184 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3185 but not after initializing Perl.
3187 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3188 output them only if debugging. By default they are written stderr,
3189 apparently, which is not desirable.
3191 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3194 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3197 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3199 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3200 authenticated or an ident call has been made. Suppress the default
3201 values for $authenticated_id and $authenticated_sender (but permit -oMai
3202 and -oMas) when testing with -bh.
3204 PH/24 Re-jigged the order of the tests in the default configuration so that the
3205 tests for valid domains and recipients precede the DNS black list and CSA
3206 tests, on the grounds that those ones are more expensive.
3208 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3209 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3210 This bug exists in every version of Exim that I still have, right back to
3213 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3214 However, an attempt to turn on case-sensitivity in a regex key by
3215 including (?-i) didn't work because the subject string was already
3216 lowercased, and the effects were non-intuitive. It turns out that a
3217 one-line patch can be used to allow (?-i) to work as expected.
3223 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3224 other effects) broke the use of negated acl sub-conditions.
3226 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3229 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3230 "Deny" causes Exim to reject the incoming connection with a 554 error.
3231 Unfortunately, if there is a major crisis, such as a disk failure,
3232 tcp-wrappers gives "deny", whereas what one would like would be some
3233 kind of temporary error. A kludge has been added to help with this.
3234 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3235 554 error is used if errno is still zero or contains ENOENT (which occurs
3236 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3239 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3241 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3242 errors. Otherwise a message that provokes a temporary error (when other
3243 messages do not) can cause a whole host to time out.
3245 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3246 addresses were routed directly to files or pipes from a redirect router.
3247 File deliveries just didn't batch; pipe deliveries might have suffered
3250 PH/06 A failure to get a lock for a hints database would erroneously always say
3251 "Failed to get write lock", even when it was really a read lock.
3253 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3254 of 0600. This has been changed to use the value of the lockfile_mode
3255 option (which defaults to 0600).
3257 PH/08 Applied small patch from the Sieve maintainer.
3259 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3260 folder from quota calculations, a direct delivery into this folder messed
3261 up the contents of the maildirsize file. This was because the regex was
3262 used only to exclude .Trash (or whatever) when the size of the mailbox
3263 was calculated. There was no check that a delivery was happening into an
3264 excluded directory. This bug has been fixed by ignoring all quota
3265 processing for deliveries into excluded directories.
3267 PH/10 Added the maildirfolder_create_regex option to appendfile.
3273 PH/01 The code for finding all the local interface addresses on a FreeBSD
3274 system running IPv6 was broken. This may well have applied to all BSD
3275 systems, as well as to others that have similar system calls. The broken
3276 code found IPv4 interfaces correctly, but gave incorrect values for the
3277 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3278 that it would not match correctly against @[] and not recognize the IPv6
3281 PH/02 The ipliteral router was not recognizing addresses of the form user@
3282 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3284 PH/03 Added disable_ipv6.
3286 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3287 decoded form, because it is most often used to construct To: headers
3288 lines in autoreplies, and the decoded form may well be syntactically
3289 invalid. However, $reply_address has leading white space removed, and all
3290 newlines turned into spaces so that the autoreply transport does not
3293 PH/05 If group was specified without a user on a router, and no group or user
3294 was specified on a transport, the group from the router was ignored.
3296 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3297 for visible compile-time settings that can be used to change these
3298 numbers, for those that want even more. Backwards compatibility with old
3299 spool files has been maintained. However, going back to a previous Exim
3300 release will lost any variables that are in spool files.
3302 PH/07 Two small changes when running in the test harness: increase delay when
3303 passing a TCP/IP connection to a new process, in case the original
3304 process has to generate a bounce, and remove special handling of
3305 127.0.0.2 (sic), which is no longer necessary.
3307 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3308 be the same on different OS.
3310 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3313 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3314 whether --show-vars was specified or not
3316 JJ/02 exipick: Added support for new ACL variable spool format introduced
3319 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3320 syntactically invalid From: or Reply-to: line, and a filter used this to
3321 generate an autoreply, and therefore failed to obtain an address for the
3322 autoreply, Exim could try to deliver to a non-existent relative file
3323 name, causing unrelated and misleading errors. What now happens is that
3324 it logs this as a hard delivery error, but does not attempt to create a
3327 PH/11 The exinext utility has a -C option for testing purposes, but although
3328 the given file was scanned by exinext itself; it wasn't being passed on
3329 when Exim was called.
3331 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3332 an end-of-file indication when reading a command response.
3334 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3335 compiled. In many other places in Exim, IPv6 addresses are always
3336 recognized, so I have changed this. It also means that IPv4 domain
3337 literals of the form [IPV4:n.n.n.n] are now always recognized.
3339 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3340 used if the router is not running as root, for example, when verifying at
3341 ACL time, or when using -bh. The debugging output from this situation was
3342 non-existent - all you got was a failure to exec. I have made two
3345 (a) Failures to set uid/gid, the current directory, or a process leader
3346 in a subprocess such as that created by queryprogram now generate
3347 suitable debugging output when -d is set.
3349 (b) The queryprogram router detects when it is not running as root,
3350 outputs suitable debugging information if -d is set, and then runs
3351 the subprocess without attempting to change uid/gid.
3353 PH/15 Minor change to Makefile for building test_host (undocumented testing
3356 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3357 additional section of a DNS packet that returns MX or SRV records.
3358 Instead, it always explicitly searches for A/AAAA records. This avoids
3359 major problems that occur when a DNS server includes only records of one
3360 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3361 fixed another bug: if SRV records were looked up and the corresponding
3362 address records were *not* found in the additional section, the port
3363 values from the SRV records were lost.
3365 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3366 using the correct key (the original address) when searching the retry
3367 rules in order to find which one to use for generating the retry hint.
3369 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3370 adding the default one. Similarly, if it contains a Reply-To: header, the
3371 errors_reply_to option, if set, is not used.
3373 PH/19 When calculating a retry time, Exim used to measure the "time since
3374 failure" by looking at the "first failed" field in the retry record. Now
3375 it does not use this if it is later than than the arrival time of the
3376 message. Instead it uses the arrival time. This makes for better
3377 behaviour in cases where some deliveries succeed, thus re-setting the
3378 "first failed" field. An example is a quota failure for a huge message
3379 when small messages continue to be delivered. Without this change, the
3380 "time since failure" will always be short, possible causing more frequent
3381 delivery attempts for the huge message than are intended.
3382 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3384 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3385 $1, $2, $3) because the numerical variables can be reset during some
3386 expansion items (e.g. "match"), thereby losing the authentication data.
3388 PH/21 Make -bV show the size of off_t variables so that the test suite can
3389 decide whether to run tests for quotas > 2G.
3391 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3392 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3393 filecount value is greater than 2G or if a quota value is greater than 2G
3394 on a system where the size of off_t is not greater than 4, a panic error
3397 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3398 never match. The debug and -bh output now contains an explicit error
3399 message indicating a malformed IPv4 address or mask.
3401 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3402 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3403 PH/23 above applies.
3405 PH/25 Do not write to syslog when running in the test harness. The only
3406 occasion when this arises is a failure to open the main or panic logs
3407 (for which there is an explicit test).
3409 PH/26 Added the /no_tell option to "control=freeze".
3411 PH/27 If a host name lookup failed very early in a connection, for example, if
3412 the IP address matched host_lookup and the reverse lookup yielded a name
3413 that did not have a forward lookup, an error message of the form "no IP
3414 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3415 could be logged. Now it outputs the IP address instead of "NULL".
3417 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3418 allows the sender and the authenticated sender to be set when
3419 submitting a message from within Exim. Since child_open_exim() is
3420 documented for local_scan(), the new function should be too.
3422 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3423 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3424 results in an empty string is now treated as unset.
3426 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3428 PH/31 Added sender_verify_fail logging option.
3430 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3431 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3432 tidied the source and removed it altogether.
3434 PH/33 When a queue run was abandoned because the load average was too high, a
3435 log line was always written; now it is written only if the queue_run log
3436 selector is set. In addition, the log line for abandonment now contains
3437 information about the queue run such as the pid. This is always present
3438 in "start" and "stop" lines but was omitted from the "abandon" line.
3440 PH/34 Omit spaces between a header name and the colon in the error message that
3441 is given when verify = headers_syntax fails (if there are lots of them,
3442 the message gets confusing).
3444 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3445 names, as there are now some PTR records that contain slashes. This check
3446 is only to protect against broken name servers that fall over on strange
3447 characters, so the fact that it applies to all lookups doesn't matter.
3449 PH/36 Now that the new test suite is complete, we can remove some of the
3450 special code in Exim that was needed for the old test suite. For example,
3451 sorting DNS records because real resolvers return them in an arbitrary
3452 order. The new test suite's fake resolver always returns records in the
3455 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3456 bounces) except when queue_only is set, to avoid logging races between
3457 the different processes.
3459 PH/38 Panic-die if .include specifies a non-absolute path.
3461 PH/39 A tweak to the "H" retry rule from its user.
3463 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3464 a label. They prevented compilation on older perls.
3466 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3467 a warning to be raised on newish perls.
3469 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3470 on queue. Changes to match documented behaviour of showing count of
3471 messages matching specified criteria.
3473 PH/40 Changed the default ident timeout from 30s to 5s.
3475 PH/41 Added support for the use of login_cap features, on those BSD systems
3476 that have them, for controlling the resources used by pipe deliveries.
3478 PH/42 The content-scanning code uses fopen() to create files in which to put
3479 message data. Previously it was not paying any attention to the mode of
3480 the files. Exim runs with umask(0) because the rest of the code creates
3481 files with open(), and sets the required mode explicitly. Thus, these
3482 files were ending up world-writeable. This was not a big issue, because,
3483 being within the spool directory, they were not world-accessible. I have
3484 created a function called modefopen, which takes an additional mode
3485 argument. It sets umask(777), creates the file, chmods it to the required
3486 mode, then resets the umask. All the relevant calls to fopen() in the
3487 content scanning code have been changed to use this function.
3489 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3490 to 24 hours. This avoids potential overflow problems when processing G
3491 and H retry rules. I suspect nobody ever tinkers with this value.
3493 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3495 PH/45 When the plaintext authenticator is running as a client, the server's
3496 challenges are checked to ensure they are valid base64 strings. By
3497 default, the authentication attempt is cancelled if an invalid string is
3498 received. Setting client_ignore_invalid_base64 true ignores these errors.
3499 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3500 they are received. Thus, the responses can be made to depend on the
3501 challenges. If an invalid string is ignored, an empty string is placed in
3504 PH/46 Messages that are created by the autoreply transport now contains a
3505 References: header, in accordance with RFCs 2822 and 3834.
3507 PH/47 Added authenticated_sender_force to the smtp transport.
3509 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3511 PH/49 Installed latest patch from the Sieve maintainer.
3513 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3514 was also set, the appendfile transport was unnecessarily scanning a
3515 directory of message files (e.g. for maildir delivery) to find the count
3516 of files (along with the size), even though it did not need this
3517 information. It now does the scan only if it needs to find either the
3518 size of the count of files.
3520 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3522 PH/52 Two bugs concerned with error handling when the smtp transport is
3525 (i) Exim was not creating retry information for temporary errors given
3526 for individual recipients after the DATA command when the smtp transport
3527 was used in LMTP mode. This meant that they could be retried too
3528 frequently, and not timed out correctly.
3530 (ii) Exim was setting the flag that allows error details to be returned
3531 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3532 recipients that were returned after the DATA command.
3534 PH/53 This is related to PH/52, but is more general: for any failing address,
3535 when detailed error information was permitted to be returned to the
3536 sender, but the error was temporary, then after the final timeout, only
3537 "retry timeout exceeded" was returned. Now it returns the full error as
3538 well as "retry timeout exceeded".
3540 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3541 do this, and (what is worse) MTAs that accept it.
3543 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3544 will now be deprecated.
3546 PH/56 New os.c-cygwin from the Cygwin maintainer.
3548 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3549 formats (previously only available in exim formats via -bpr, -bpru,
3550 and -bpra. Now also available in native and exiqgrep formats)
3552 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3553 with very large, slow to parse queues
3555 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3557 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3559 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3560 responses to authentication challenges, though it was showing the
3561 challenges; (ii) I've removed the CR characters from the debug output for
3564 PH/58 Allow for the insertion of a newline as well as a space when a string
3565 is turned into more than one encoded-word during RFC 2047 encoding. The
3566 Sieve code now uses this.
3568 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3569 data_4xx, lost_connection, tls_required.
3571 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3572 message was being sent as an SMTP response.
3574 PH/61 Add -l and -k options to exicyclog.
3576 PH/62 When verifying, if an address was redirected to one new address, so that
3577 verification continued, and the new address failed or deferred after
3578 having set something in $address_data, the value of $address_data was not
3579 passed back to the ACL. This was different to the case when no
3580 redirection occurred. The value is now passed back in both cases.
3582 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3583 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3584 use login_cap.h, so on its own it isn't the distinguishing feature. The
3585 new name refers directly to the setclassresources() function.
3587 PH/65 Added configuration files for NetBSD3.
3589 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3591 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3592 is preferred over IPv4.
3594 PH/68 The bounce_return_message and bounce_return_body options were not being
3595 honoured for bounces generated during the reception of non-SMTP messages.
3596 In particular, this applied to messages rejected by the ACL. This bug has
3597 been fixed. However, if bounce_return_message is true and bounce_return_
3598 body is false, the headers that are returned for a non-SMTP message
3599 include only those that have been read before the error was detected.
3600 (In the case of an ACL rejection, they have all been read.)
3602 PH/69 The HTML version of the specification is now built in a directory called
3603 spec_html instead of spec.html, because the latter looks like a path with
3604 a MIME-type, and this confuses some software.
3606 PH/70 Catch two compiler warnings in sieve.c.
3608 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3609 function verify_get_ident() calls ip_connect() to connect a socket, but
3610 if the "connect()" function timed out, ip_connect() used to close the
3611 socket. However, verify_get_ident() also closes the socket later, and in
3612 between Exim writes to the log, which may get opened at this point. When
3613 the socket was closed in ip_connect(), the log could get the same file
3614 descriptor number as the socket. This naturally causes chaos. The fix is
3615 not to close the socket in ip_connect(); the socket should be closed by
3616 the function that creates it. There was only one place in the code where
3617 this was missing, in the iplookup router, which I don't think anybody now
3618 uses, but I've fixed it anyway.
3620 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3621 well as to direct DNS lookups. Otherwise the handling of names in host
3622 lists is inconsistent and therefore confusing.
3628 PH/01 Two changes to the default runtime configuration:
3630 (1) Move the checks for relay_from_hosts and authenticated clients from
3631 after to before the (commented out) DNS black list checks.
3633 (2) Add control=submission to the relay_from_hosts and authenticated
3634 clients checks, on the grounds that messages accepted by these
3635 statements are most likely to be submissions.
3637 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3639 (1) Generate an error if the third argument for the ${prvs expansion is
3642 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3645 (3) Reset the variables that are obtained from the first argument of
3646 ${prvscheck and used in the second argument before leaving the code,
3647 because their memory is reclaimed, so using them afterwards may do
3650 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3651 one (it's much easier than Tom thought :-).
3653 (5) Because of (4), we can now allow for the use of $prvscheck_result
3654 inside the third argument.
3656 PH/03 For some reason, the default setting of PATH when running a command from
3657 a pipe transport was just "/usr/bin". I have changed it to
3660 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3661 anything to be listed in the output from -bV.
3663 PH/05 When a filter generated an autoreply, the entire To: header line was
3664 quoted in the delivery log line, like this:
3666 => >A.N.Other <ano@some.domain> <original@ddress> ...
3668 This has been changed so that it extracts the operative address. There
3669 may be more than one such address. If so, they are comma-separated, like
3672 => >ano@some.domain,ona@other.domain <original@ddress> ...
3674 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3675 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3676 not being looked up in the rDNS to get a host name, Exim was showing the
3677 IP address twice in Received: lines, even though the IP addresses were
3678 identical. For example:
3680 Received: from [1.2.3.4] (helo=[1.2.3.4])
3682 However, if the real host name was known, it was omitting the HELO data
3683 if it matched the actual IP address. This has been tidied up so that it
3684 doesn't show the same IP address twice.
3686 PH/07 When both +timestamp and +memory debugging was on, the value given by
3687 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3688 called by the debug printing, thereby overwriting the timestamp buffer.
3689 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3691 PH/08 When the original message was included in an autoreply transport, it
3692 always said "this is a copy of the message, including all the headers",
3693 even if body_only or headers_only was set. It now gives an appropriate
3696 PH/09 Applied a patch from the Sieve maintainer which:
3698 o fixes some comments
3699 o adds the (disabled) notify extension core
3700 o adds some debug output for the result of if/elsif tests
3701 o points to the current vacation draft in the documentation
3702 and documents the missing references header update
3706 o fixes a bug in processing the envelope test (when testing
3707 multiple envelope elements, the last element determined the
3710 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3711 Electronic Mail") by including:
3713 Auto-submitted: auto-generated
3715 in the messages that it generates (bounce messages and others, such as
3716 warnings). In the case of bounce messages for non-SMTP messages, there was
3717 also a typo: it was using "Auto_submitted" (underscore instead of
3718 hyphen). Since every message generated by Exim is necessarily in response
3719 to another message, thes have all been changed to:
3721 Auto-Submitted: auto-replied
3723 in accordance with these statements in the RFC:
3725 The auto-replied keyword:
3727 - SHOULD be used on messages sent in direct response to another
3728 message by an automatic process,
3730 - MUST NOT be used on manually-generated messages,
3732 - MAY be used on Delivery Status Notifications (DSNs) and Message
3733 Disposition Notifications (MDNs),
3735 - MUST NOT be used on messages generated by automatic or periodic
3736 processes, except for messages which are automatic responses to
3739 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3740 to the default Received: header definition.
3742 PH/12 Added log selector acl_warn_skipped (default on).
3744 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3745 variables because (a) they are in the wrong storage pool and (b) even if
3746 they were copied, it wouldn't work properly because of the caching.
3748 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3749 checking when decoding. Apparently there are clients that generate
3750 overlong encoded strings. Why am I not surprised?
3752 PH/15 If the first argument of "${if match_address" was not empty, but did not
3753 contain an "@" character, Exim crashed. Now it writes a panic log message
3754 and treats the condition as false.
3756 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3758 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3759 extension "envelope-auth". The code is finished and in agreement with
3760 other implementations, but there is no documentation so far and in fact,
3761 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3762 not changing the active code.
3764 Print executed "if" and "elsif" statements when debugging is used. This
3765 helps a great deal to understand what a filter does.
3767 Document more things not specified clearly in RFC3028. I had all this
3768 sorted out, when out of a sudden new issues came to my mind. Oops."
3770 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3773 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3774 canonical form (as documented). However, after a host name lookup from
3775 the IP address, check_host() was doing a simple string comparison with
3776 addresses acquired from the DNS when checking that the found name did
3777 have the original IP as one of its addresses. Since any found IPv6
3778 addresses are likely to be in abbreviated form, the comparison could
3779 fail. Luckily, there already exists a function for doing the comparison
3780 by converting both addresses to binary, so now that is used instead of
3781 the text comparison.
3783 PH/20 There was another similar case to PH/19, when a complete host name was
3784 given in a host list; looking up its IP address could give an abbreviated
3785 form, whereas the current host's name might or might not be abbreviated.
3786 The same fix has been applied.
3792 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3793 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3796 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3797 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3799 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3801 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3802 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3803 Also, for "Auto-Submitted", treat anything other than "no" as
3804 non-personal, in accordance with RFC 3834. (Previously it treated
3805 anything starting "auto-" as non-personal.)
3807 TF/01 The control=submission/name=... option had a problem with syntax
3808 errors if the name included a slash character. The /name= option
3809 now slurps the rest of the string, so it can include any characters
3810 but it must come last in the list of options (after /sender_retain
3813 PH/05 Some modifications to the interface to the fake nameserver for the new
3821 TK/01 Added the "success_on_redirect" address verification option. See
3822 NewStuff for rationale and an example.
3824 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3826 PH/02 Patch to exigrep to allow it to work on syslog lines.
3828 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3829 fread() to skip over the body file's header line, because in Cygwin the
3830 header line is locked and is inaccessible.
3832 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3833 co-exist for some time) to make it clear that it is the Exim ID that is
3834 referenced, not the Message-ID: header line.
3836 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3837 string_format() function, because snprintf() does not exist on all
3840 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3841 ${stat: expansion item.
3843 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3844 protocol synchronization error", to keep the pedants happy.
3846 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3847 well as for IRIX systems, when gcc is being used. See the host.c source
3850 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3852 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3855 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3856 transport and to the smtp transport in LMTP mode.
3858 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3860 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3861 run a filter in a subprocess. This could lead to confusion in subsequent
3862 lookups in the parent process. There should also be a search_tidyup() at
3863 the end of the subprocess.
3865 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3866 only if the host matched helo_try_verify_hosts, which caused the
3867 verification to occur when the EHLO/HELO command was issued. The ACL just
3868 tested the remembered result. Now, if a previous verification attempt has
3869 not happened, "verify = helo" does it there and then.
3871 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3873 TK/03 Fix log output including CR from clamd.
3875 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3876 exist provoked a memory error which could cause a segfault.
3878 PH/15 Installed PCRE 6.2
3880 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3882 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3883 of the problem. Specifically, suggested +O2 rather than +O1 for the
3886 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3888 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3889 record, the port was not correctly logged when the outgoing_port log
3890 selector was set (it logged the transort's default port).
3892 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3893 fallback_hosts, and "hosts" in the smtp transport.
3895 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3896 host errors such as "Connection refused".
3898 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3899 authentication with radiusclient 0.4.9:
3901 - Error returned from rc_read_config was caught wrongly
3902 - Username/password not passed on to radius server due to wrong length.
3904 The presumption is that some radiusclient API changes for 4.51/PH/17
3905 were not taken care of correctly. The code is still untested by me (my
3906 Linux distribution still has 0.3.2 of radiusclient), but it was
3907 contributed by a Radius user.
3909 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3910 expanding the "port" option of the smtp transport.
3912 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3913 while reading a MIME header. Thanks to Tom Hughes for a patch.
3915 PH/24 Include config.h inside local_scan.h so that configuration settings are
3918 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3919 that in an ACL for RCPT (for example), you can examine exactly what was
3922 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3923 commands, but it was not correctly comparing the address with the actual
3924 client host address. Thus, it would show the EHLO address in Received:
3925 header lines when this was not necessary.
3927 PH/27 Added the % operator to ${eval:}.
3929 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3930 it should be ignoring failures (because with -C, for example, it has lost
3931 privilege). It wasn't ignoring creation failures other than "already
3934 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3937 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3938 because an input line was too long, either on its own, or by virtue of
3939 too many continuations, the temporary file was not being removed, and the
3940 return code was incorrect.
3942 PH/31 Missing "BOOL" in function definition in filtertest.c.
3944 PH/32 Applied Sieve patches from the maintainer.
3946 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3948 PH/33 Added "verify = not_blind".
3950 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3951 Local/Makefile (with some defaults set). These are used in built scripts
3952 such as exicyclog, but they have never been used in the exim_install
3953 script (though there are many overriding facilities there). I have
3954 arranged that the exim_install script now takes note of these two
3957 PH/35 Installed configuration files for Dragonfly.
3959 PH/36 When a locally submitted message by a trusted user did not contain a
3960 From: header, and the sender address was obtained from -f or from an SMTP
3961 MAIL command, and the trusted user did not use -F to supply a sender
3962 name, $originator_name was incorrectly used when constructing a From:
3963 header. Furthermore, $originator_name was used for submission mode
3964 messages from external hosts without From: headers in a similar way,
3965 which is clearly wrong.
3967 PH/37 Added control=suppress_local_fixups.
3969 PH/38 When log_selector = +received_sender was set, and the addition of the
3970 sender made the log line's construction buffer exactly full, or one byte
3971 less than full, an overflow happened when the terminating "\n" was
3974 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3975 when the result of a list match is failure because a DNS lookup failed.
3977 PH/40 RM_COMMAND is now used in the building process.
3979 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3980 the "build-* directories that it finds.
3982 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3983 address in a domain literal was a prefix of an interface address.
3985 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3986 when verifying a sender address, unless rewrite_headers is false.
3988 PH/44 Wrote a long comment about why errors_to addresses are verified as
3989 recipients, not senders.
3991 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3992 the ratelimit ACL was added.
3994 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3996 PH/46 Added extra information about PostgreSQL errors to the error string.
3998 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3999 avoiding the need to install special zones in a real server. This is
4000 backwards compatible; if it can't find the fake resolver, it drops back.
4001 Thus, both old and new test suites can be run.
4003 TF/02 Added util/ratelimit.pl
4005 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4006 clock is set back in time.
4008 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4009 Candler <B.Candler@pobox.com>.
4011 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4012 OK for addresses that are the result of redirections.
4014 PH/48 A number of further additions for the benefit of the new test suite,
4015 including a fake gethostbyname() that interfaces to the fake DNS resolver
4018 TF/06 The fix for widen_domains has also been applied to qualify_single and
4019 search_parents which are the other dnslookup options that can cause
4022 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4025 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4027 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4028 service" if it encountered an MX record with an empty target hostname.
4029 The message is now "an MX or SRV record indicated no SMTP service".
4031 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4032 if the DNS of the sending site is misconfigured. This is quite a
4033 common situation. This change restores the behaviour of treating a
4034 helo verification defer as a failure.
4036 PH/51 If self=fail was set on a router, the bounce message did not include the
4037 actual error message.
4043 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4045 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4046 command itself ended in error, the underlying message about the transport
4047 filter timeout was being overwritten with the pipe command error. Now the
4048 underlying error message should be appended to the second error message.
4050 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4052 PH/02 Reduce the amount of output that "make" produces by default. Full output
4053 can still be requested.
4055 PH/03 The warning log line about a condition test deferring for a "warn" verb
4056 was being output only once per connection, rather than after each
4057 occurrence (because it was using the same function as for successful
4058 "warn" verbs). This seems wrong, so I have changed it.
4060 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4061 it should not have, which might have caused a crash in the right
4062 circumstances, but probably never did.
4064 PH/04 Installed a modified version of Tony Finch's patch to make submission
4065 mode fix the return path as well as the Sender: header line, and to
4066 add a /name= option so that you can make the user's friendly name appear
4069 TF/03 Added the control = fakedefer ACL modifier.
4071 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4072 Mark Lowes for thorough testing.
4074 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4076 TK/03 Merged latest SRS patch from Miles Wilton.
4078 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4079 with the definition in sysexits.h (which is #included earlier).
4080 Fortunately, Exim does not actually use EX_OK. The code used to try to
4081 preserve the sysexits.h value, by assuming that macro definitions were
4082 scanned for macro replacements. I have been disabused of this notion,
4083 so now the code just undefines EX_OK before #including unistd.h.
4085 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4086 in the smtp transport. When a block could not be written in a single
4087 write() function, the timeout was being re-applied to each part-write.
4088 This seems wrong - if the receiver was accepting one byte at a time it
4089 would take for ever. The timeout is now adjusted when this happens. It
4090 doesn't have to be particularly precise.
4092 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4093 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4095 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4096 without a domain if the check to <postmaster@domain> fails.
4098 SC/01 Eximstats: added -xls and the ability to specify output files
4099 (patch written by Frank Heydlauf).
4101 SC/02 Eximstats: use FileHandles for outputting results.
4103 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4105 SC/04 Eximstats: fixed display of large numbers with -nvr option
4107 SC/05 Eximstats: fixed merging of reports with empty tables.
4109 SC/06 Eximstats: added the -include_original_destination flag
4111 SC/07 Eximstats: removed tabs and trailing whitespace.
4113 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4115 TK/06 MBOX spool code: Add real "From " MBOX separator line
4116 so the .eml file is really in mbox format (even though
4117 most programs do not really care). Patch from Alex Miller.
4119 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4120 The latter is generated from $received_to and is only set if the
4121 message has one envelope recipient. SA can use these headers,
4122 obviously out-of-the-box. Patch from Alex Miller.
4124 PH/08 The ${def test on a variable was returning false if the variable's
4125 value was "0", contrary to what the specification has always said!
4126 The result should be true unless the variable is empty.
4128 PH/09 The syntax error of a character other than { following "${if
4129 def:variable_name" (after optional whitespace) was not being diagnosed.
4130 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4131 accidental colon was present, for example, could give incorrect results.
4133 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4134 result is used (not including appendfile, where other changes are about
4137 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4138 This involved changing a lot of size variables from int to off_t. It
4139 should work with maildirs and everything.
4141 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4142 spamd dying while we are connected to it.
4144 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4147 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4148 function for BDB 4.3.
4150 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4152 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4153 the sources for PCRE have been re-arranged and more files are now
4156 PH/15 The code I had for printing potentially long long variables in PH/11
4157 above was not the best (it lost precision). The length of off_t variables
4158 is now inspected at build time, and an appropriate printing format (%ld
4159 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4160 to be "long long int" or "long int". This is needed for the internal
4161 formatting function string_vformat().
4163 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4164 the configuration file to be ":syslog", then the script "guesses" where
4165 the logs files are, rather than using the compiled in default. In our
4166 case the guess is not the same as the compiled default, so the script
4167 suddenly stopped working when I started to use syslog. The patch checks
4168 to see if log_file_path is "". If so, it attempts to read it from exim
4169 with no configuration file to get the compiled in version, before it
4170 falls back to the previous guessing code."
4172 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4173 implementing BATV in an Exim configuration. See NewStuff for the gory
4176 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4177 Makefile that are specific to HP-UX.
4179 PH/18 If the "use_postmaster" option was set for a recipient callout together
4180 with the "random" option, the postmaster address was used as the MAIL
4181 FROM address for the random test, but not for the subsequent recipient
4182 test. It is now used for both.
4184 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4185 patch removes a few documentation additions to RFC 3028, because the
4186 latest draft now contains them. It adds the new en;ascii-case comparator
4187 and a new error check for 8bit text in MIME parts. Comparator and
4188 require names are now matched exactly. I enabled the subaddress
4189 extension, but it is not well tested yet (read: it works for me)."
4191 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4192 rework some of the code of TK/09 above to avoid the hardwired use of
4193 "%lld" and "long long". Replaced the call to snprintf() with a call to
4196 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4197 records point to non-existent hosts", "retry timeout exceeded", and
4198 "retry time not reached for any host after a long failure period".
4200 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4201 experimental DomainKeys support:
4203 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4204 (2) On an error such as an illegally used "control", the wrong name for
4205 the control was given.
4207 These problems did NOT occur unless DomainKeys support was compiled.
4209 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4211 PH/24 Added ${if match_ip condition.
4213 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4214 databases so that it will be absolutely obvious if a crash occurs in the
4215 DB library. This is a regular occurrence (often caused by mis-matched
4218 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4219 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4220 picked up on a user's system that detects such things. There doesn't seem
4221 to be a gcc warning option for this - only an attribute that has to be
4222 put on the function's prototype. It seems that in Fedora Core 4 they have
4223 set this on a number of new functions. No doubt there will be more in due
4226 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4227 specify a transport. However, if an address that was verified by such a
4228 router was the subject of a callout, Exim crashed because it tried to
4229 read the rcpt_include_affixes from the non-existent transport. Now it
4230 just assumes that the setting of that option is false. This bug was
4231 introduced by 4.51/PH/31.
4233 PH/28 Changed -d+all to exclude +memory, because that information is very
4234 rarely of interest, but it makes the output a lot bigger. People tend to
4235 do -d+all out of habit.
4237 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4238 code in os-type was giving problems when libc.so lives in lib64, like on
4241 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4242 aren't the modern standard, and it seems that some systems' include files
4243 don't always have them. Exim was already checking for some of the newer
4244 ones like T_AAAA, and defining it itself. I've added checks for all the
4245 record types that Exim uses.
4247 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4248 not automatically generating a new one, as it is supposed to. This
4249 prevented TLS from working. If the file did exist, but contained invalid
4250 data, a new version was generated, as expected. It was only the case of a
4251 non-existent file that was broken.
4253 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4254 with a change in libdomainkeys > 0.64.
4256 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4257 from DNS. If the selector record carries the flag, it now has
4258 precedence over the domain-wide flag.
4260 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4262 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4263 the use of a port name that isn't defined in /etc/services) occurred, the
4264 message was deferred as in a normal delivery, and thus remained on the
4265 spool, instead of being failed because of the mua_wrapper setting. This
4266 is now fixed, and I tidied up some of the mua_wrapper messages at the
4269 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4270 the messages in a hash of arrays rather than using individual hashes.
4271 This is a bit cleaner and results in dramatic memory savings, albeit
4272 at a slight CPU cost.
4274 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4275 as requested by Marc Sherman.
4277 SC/10 Eximstats: added histograms for user specified patterns as requested
4280 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4282 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4283 fopen() in the content-scanning modules that did not already have it.
4289 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4290 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4292 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4294 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4296 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4297 to test Sieve filters that use "vacation".
4299 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4300 that changes the way the GnuTLS parameters are stored in the cache file.
4301 The new format can be generated externally. For backward compatibility,
4302 if the data in the cache doesn't make sense, Exim assumes it has read an
4303 old-format file, and it generates new data and writes a new file. This
4304 means that you can't go back to an older release without removing the
4307 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4308 work if there are any delivery delays because "one_time" forces the
4309 parent to be marked "delivered", so its unseen clone is never tried
4310 again. For this reason, Exim now forbids the simultaneous setting of
4313 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4314 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4315 entry if you want to know the details. The fix, however, neglected to
4316 consider the case when local delivery batching is involved. The test for
4317 "previously delivered" was not happening when checking to see if an
4318 address could be batched with a previous (undelivered) one; under
4319 certain circumstances this could lead to multiple deliveries to the same
4322 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4323 in its include files, and this causes problems building Exim.
4325 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4326 header_syntax) but Exim was just ignoring anything given after a slash.
4327 In particular, this caused confusion with an attempt to use "verify =
4328 reverse_host_lookup/defer_ok". An error is now given when options are
4329 supplied for verify items that do not have them. (Maybe reverse_host_
4330 lookup should have a defer_ok option, but that's a different point.)
4332 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4333 defined by RFC 821) to 2048, because there were problems with some AUTH
4334 commands, and RFC 1869 says the size should be increased for extended
4335 SMTP commands that take arguments.
4337 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4340 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4341 "unknown" error; now it says that the functionality isn't in the binary.
4343 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4344 an address' error message when a string expansion fails (syntax or
4345 whatever). Otherwise the password may appear in the log. Following change
4346 PH/42 below, there is no longer a chance of it appearing in a bounce
4349 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4351 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4352 of its addresses was ever tried. (Bugzilla bug #2.)
4354 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4355 the result incorrectly in the debug output. (It correctly added a newline
4356 to what was transported.)
4358 TF/01 Added $received_time.
4360 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4361 commented out examples of how to interface to a virus scanner and to
4362 SpamAssassin. Also added commented examples of av_scanner and
4363 spamd_address settings.
4365 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4366 and controls are allowed in which ACLs. There were a couple of minor
4367 errors. Some of the entries in the conditions table (which is a table of
4368 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4369 negation of where the condition IS allowed.
4371 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4373 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4374 header file does not have a version number, so I've had to invent a new
4375 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4376 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4377 radiusclient), but it was contributed by a Radius user.
4379 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4380 files or directories, for OpenSSL.
4382 PH/19 When an Exim process that is running as root has to create an Exim log
4383 file, it does so in a subprocess that runs as exim:exim so as to get the
4384 ownership right at creation (otherwise, other Exim processes might see
4385 the file with the wrong ownership). There was no test for failure of this
4386 fork() call, which would lead to the process getting stuck as it waited
4387 for a non-existent subprocess. Forks do occasionally fail when resources
4388 run out. I reviewed all the other calls to fork(); they all seem to check
4391 PH/20 When checking for unexpected SMTP input at connect time (before writing
4392 the banner), Exim was not dealing correctly with a non-positive return
4393 from the read() function. If the client had disconnected by this time,
4394 the result was a log entry for a synchronization error with an empty
4395 string after "input=" when read() returned zero. If read() returned -1
4396 (an event I could not check), uninitialized data bytes were printed.
4397 There were reports of junk text (parts of files, etc) appearing after
4400 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4402 PH/22 Added support for macro redefinition, and (re)definition in between
4403 driver and ACL definitions.
4405 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4406 forgetting to use the resulting value; it was using the unexpanded value.
4408 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4409 hadn't been configured. The fix is from Juergen Kreileder, who
4410 understands it better than I do:
4412 "Here's what I see happening with three configured cyrus_sasl
4413 authenticators configured (plain, login, cram-md5):
4415 On startup auth_cyrus_sasl_init() gets called for each of these.
4416 This means three calls to sasl_listmech() without a specified mech_list.
4417 => SASL tests which mechs of all available mechs actually work
4418 => three warnings about OTP not working
4419 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4421 With the patch, sasl_listmech() also gets called three times. But now
4422 SASL's mech_list option is set to the server_mech specified in the the
4423 authenticator. Or in other words, the answer from sasl_listmech()
4424 gets limited to just the mech you're testing for (which is different
4426 => the return list contains just 'plain' or 'login', 'cram-md5' or
4427 nothing depending on the value of ob->server_mech.
4429 I've just tested the patch: Authentication still works fine,
4430 unavailable mechs specified in the exim configuration are still
4431 caught, and the auth.log warnings about OTP are gone."
4433 PH/25 When debugging is enabled, the contents of the command line are added
4434 to the debugging output, even when log_selector=+arguments is not
4437 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4438 answer is "GNU", and only if the return is "GNU/something" is the answer
4441 PH/27 $acl_verify_message is now set immediately after the failure of a
4442 verification in an ACL, and so is available in subsequent modifiers. In
4443 particular, the message can be preserved by coding like this:
4445 warn !verify = sender
4446 set acl_m0 = $acl_verify_message
4448 Previously, $acl_verify_message was set only while expanding "message"
4449 and "log_message" when a very denied access.
4451 PH/28 Modified OS/os.c-Linux with
4453 -#ifndef OS_LOAD_AVERAGE
4454 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4456 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4457 nomenclature these days.)
4459 PH/29 Installed patch from the Sieve maintainer that adds the options
4460 sieve_useraddress and sieve_subaddress to the redirect router.
4462 PH/30 In these circumstances:
4463 . Two addresses routed to the same list of hosts;
4464 . First host does not offer TLS;
4465 . First host accepts first address;
4466 . First host gives temporary error to second address;
4467 . Second host offers TLS and a TLS session is established;
4468 . Second host accepts second address.
4469 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4470 and peerdn, if requested) that were in fact used only for the second
4473 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4474 attention to any local part prefix or suffix that was matched by the
4475 router that accepted the address. It now behaves in the same way as it
4476 does for delivery: the affixes are removed from the local part unless
4477 rcpt_include_affixes is set on the transport.
4479 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4480 timeout during the DATA phase of an incoming message.
4482 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4483 applied a patch sanctioned by the Sieve maintainer.
4485 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4486 the uid or gid is negative. A case of a negative gid caused this to be
4487 noticed. The fix allows for either to be negative.
4489 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4490 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4493 PH/36 The change PH/12 above was broken. Fixed it.
4495 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4496 the grounds that routing the same address twice would always produce the
4497 same answer. This might have been true once, but it is certainly no
4498 longer true now. Routing a child address may depend on the previous
4499 routing that produced that child. Some complicated redirection strategies
4500 went wrong when messages had multiple recipients, and made Exim's
4501 behaviour dependent on the order in which the addresses were given.
4503 I have moved the duplicate checking until after the routing is complete.
4504 Exim scans the addresses that are assigned to local and remote
4505 transports, and removes any duplicates. This means that more work will be
4506 done, as duplicates will always all be routed, but duplicates are
4507 presumably rare, so I don't expect this is of any significance.
4509 For deliveries to pipes, files, and autoreplies, the duplicate checking
4510 still happens during the routing process, since they are not going to be
4513 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4514 It corrects a timeout issue with spamd. This is Ian's comment: "The
4515 background is that sometimes spamd either never reads data from a
4516 connection it has accepted, or it never writes response data. The exiscan
4517 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4518 blindly assumes that writes won't block so it may never time out."
4520 PH/39 Allow G after quota size as well as K and M.
4522 PH/40 The value set for $authenticated_id in an authenticator may not contain
4523 binary zeroes or newlines because the value is written to log lines and
4524 to spool files. There was no check on this. Now the value is run through
4525 the string_printing() function so that such characters are converted to
4526 printable escape sequences.
4528 PH/41 $message_linecount is a new variable that contains the total number of
4529 lines in the message. Compare $body_linecount, which is the count for the
4532 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4533 bounce and delay warning messages, except in certain special cases, which
4536 (a) An SMTP error message from a remote host;
4537 (b) A message specified in a :fail: redirection;
4538 (c) A message specified in a "fail" command in a system filter;
4539 (d) A message specified in a FAIL return from the queryprogram router;
4540 (e) A message specified by the cannot_route_message router option.
4542 In these cases only, Exim does include the error details in bounce and
4543 warning messages. There are also a few cases where bland messages such
4544 as "unrouteable address" or "local delivery error" are given.
4546 PH/43 $value is now also set for the "else" part of a ${run expansion.
4548 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4549 being worked on, but at least Exim now implements the latest version to
4552 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4553 process to complete was treated as a delivery failure, a timeout while
4554 writing the message to the pipe was logged, but erroneously treated as a
4555 successful delivery. Such timeouts include transport filter timeouts. For
4556 consistency with the overall process timeout, these timeouts are now
4557 treated as errors, giving rise to delivery failures by default. However,
4558 there is now a new Boolean option for the pipe transport called
4559 timeout_defer, which, if set TRUE, converts the failures into defers for
4560 both kinds of timeout. A transport filter timeout is now identified in
4563 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4564 systems where "make" and "gmake" are different, calling "gmake" at top
4565 level broke things. I've arranged for the value of $(MAKE) to be passed
4566 from the Makefile to this script so that it can call the same version of
4570 A note about Exim versions 4.44 and 4.50
4571 ----------------------------------------
4573 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4574 changes of various kinds. As a consequence, a big documentation update was
4575 needed. This delayed the release for rather longer than seemed good, especially
4576 in the light of a couple of (minor) security issues. Therefore, the changes
4577 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4578 release. So 4.44 and 4.50 are in effect two different branches that both start
4581 I have left the 4.50 change log unchanged; it contains all the changes since
4582 4.43. The change log for 4.44 is below; many of its items are identical to
4583 those for 4.50. This seems to be the most sensible way to preserve the
4584 historical information.
4590 1. Minor wording change to the doc/README.SIEVE file.
4592 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4593 computation of the current number of files was incorrect.
4595 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4596 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4599 4. Give more explanation in the error message when the command for a transport
4600 filter fails to execute.
4602 5. There are several places where Exim runs a non-Exim command in a
4603 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4604 was being done only for the command run by the queryprogram router. It is
4605 now done for all such subprocesses. The other cases are: ${run, transport
4606 filters, and the commands run by the lmtp and pipe transports.
4608 6. Added CONFIGURE_GROUP build-time option.
4610 7. Some older OS have a limit of 256 on the maximum number of file
4611 descriptors. Exim was using setrlimit() to set 1000 as a large value
4612 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4613 systems. I've change it so that if it can't get 1000, it tries for 256.
4615 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4616 was an oversight, and furthermore, ever since the addition of extra
4617 controls (e.g. 4.43/32), the checks on when to allow different forms of
4618 "control" were broken. There should now be diagnostics for all cases when a
4619 control that does not make sense is encountered.
4621 9. Added the /retain_sender option to "control=submission".
4623 10. $recipients is now available in the predata ACL (oversight).
4625 11. Tidy the search cache before the fork to do a delivery from a message
4626 received from the command line. Otherwise the child will trigger a lookup
4627 failure and thereby defer the delivery if it tries to use (for example) a
4628 cached ldap connection that the parent has called unbind on.
4630 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4631 of $address_data from the recipient verification was clobbered by the
4632 sender verification.
4634 13. The value of address_data from a sender verification is now available in
4635 $sender_address_data in subsequent conditions in the ACL statement.
4637 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4639 15. Added a new option "connect=<time>" to callout options, to set a different
4642 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4643 was its contents. (It was OK if the option was not defined at all.)
4645 17. A "Completed" log line is now written for messages that are removed from
4646 the spool by the -Mrm option.
4648 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4649 information about exactly what failed.
4651 19. Added -dd to debug only the daemon process.
4653 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4654 handles timeouts, both on the server side and network timeouts. Renamed the
4655 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4657 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4658 It is now set to "smtps".
4660 22. $host_address is now set to the target address during the checking of
4661 ignore_target_hosts.
4663 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4664 being passed; this would have caused $sender_host_name to have been used if
4665 matching the list had actually called for a host name (not very likely,
4666 since this list is usually IP addresses). A host name is now passed as
4669 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4670 code that specifies a non-restarting handler (typically sigaction() in
4671 modern systems) in an attempt to fix a rare and obscure crash bug.
4673 25. Narrowed the window for a race in the daemon that could cause it to ignore
4674 SIGCHLD signals. This is not a major problem, because they are used only to
4675 wake it up if nothing else does.
4677 26. A malformed maildirsize file could cause Exim to calculate negative values
4678 for the mailbox size or file count. Odd effects could occur as a result.
4679 The maildirsize information is now recalculated if the size or filecount
4682 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4683 support for a long time. Removed HAVE_SYS_VFS_H.
4685 28. Installed the latest version of exipick from John Jetmore.
4687 29. In an address list, if the pattern was not a regular expression, an empty
4688 subject address (from a bounce message) matched only if the pattern was an
4689 empty string. Non-empty patterns were not even tested. This was the wrong
4690 because it is perfectly reasonable to use an empty address as part of a
4691 database query. An empty address is now tested by patterns that are
4692 lookups. However, all the other forms of pattern expect the subject to
4693 contain a local part and a domain, and therefore, for them, an empty
4694 address still always fails if the pattern is not itself empty.
4696 30. Exim went into a mad DNS loop when attempting to do a callout where the
4697 host was specified on an smtp transport, and looking it up yielded more
4698 than one IP address.
4700 31. Re-factored the code for checking spool and log partition space into a
4701 function that finds that data and another that does the check. The former
4702 is then used to implement four new variables: $spool_space, $log_space,
4703 $spool_inodes, and $log_inodes.
4705 32. The RFC2047 encoding function was originally intended for short strings
4706 such as real names; it was not keeping to the 75-character limit for
4707 encoded words that the RFC imposes. It now respects the limit, and
4708 generates multiple encoded words if necessary. To be on the safe side, I
4709 have increased the buffer size for the ${rfc2047: expansion operator from
4712 33. It is now permitted to omit both strings after an "if" condition; if the
4713 condition is true, the result is "true". As before, when the second string
4714 is omitted, a false condition yields an empty string. This makes it less
4715 cumbersome to write custom ACL and router conditions.
4717 34. Failure to deliver a bounce message always caused it to be frozen, even if
4718 there was an errors_to setting on the router. The errors_to setting is now
4721 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4722 canonical form (fully expanded) before being placed in
4723 $sender_host_address.
4725 36. The table in the code that translates DNS record types into text (T_A to
4726 "A" for instance) was missing entries for NS and CNAME. It is just possible
4727 that this could have caused confusion if both these types were looked up
4728 for the same domain, because the text type is used as part of Exim's
4729 per-process caching. But the chance of anyone hitting this buglet seems
4732 37. The dnsdb lookup has been extended in a number of ways.
4734 (1) There is a new type, "zns", which walks up the domain tree until it
4735 finds some nameserver records. It should be used with care.
4737 (2) There is a new type, "mxh", which is like "mx" except that it returns
4738 just the host names, not the priorities.
4740 (3) It is now possible to give a list of domains (or IP addresses) to be
4741 looked up. The behaviour when one of the lookups defers can be
4742 controlled by a keyword.
4744 (4) It is now possible to specify the separator character for use when
4745 multiple records are returned.
4747 38. The dnslists ACL condition has been extended: it is now possible to supply
4748 a list of IP addresses and/or domains to be looked up in a particular DNS
4751 39. Added log_selector=+queue_time_overall.
4753 40. When running the queue in the test harness, wait just a tad after forking a
4754 delivery process, to get repeatability of debugging output.
4756 41. Include certificate and key file names in error message when GnuTLS fails
4757 to set them up, because the GnuTLS error message doesn't include the name
4758 of the failing file when there is a problem reading it.
4760 42. Allow both -bf and -bF in the same test run.
4762 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4764 44. The "Exiscan patch" is now merged into the mainline Exim source.
4766 45. Sometimes the final signoff response after QUIT could fail to get
4767 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4768 < 0 before doing a fflush(). This bug looks as though it goes back to the
4769 introduction of TLS in release 3.20, but "sometimes" must have been rare
4770 because the tests only now provoked it.
4772 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4773 (this can affect the format of dates).
4775 47. exim_tidydb, when checking for the continued existence of a message for
4776 which it has found a message-specific retry record, was not finding
4777 messages that were in split spool directories. Consequently, it was
4778 deleting retry records that should have stayed in existence.
4780 48. Steve fixed some bugs in eximstats.
4782 49. The SPA authentication driver was not abandoning authentication and moving
4783 on to the next authenticator when an expansion was forced to fail,
4784 contradicting the general specification for all authenticators. Instead it
4785 was generating a temporary error. It now behaves as specified.
4787 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4788 (the order specifies the preference for clients). The order is now AES256,
4789 AES128, 3DES, ARCFOUR128.
4791 51. Small patch to Sieve code - explicitly set From: when generating an
4794 52. Exim crashed if a remote delivery caused a very long error message to be
4795 recorded - for instance if somebody sent an entire SpamAssassin report back
4796 as a large number of 550 error lines. This bug was coincidentally fixed by
4797 increasing the size of one of Exim's internal buffers (big_buffer) that
4798 happened as part of the Exiscan merge. However, to be on the safe side, I
4799 have made the code more robust (and fixed the comments that describe what
4802 53. Now that there can be additional text after "Completed" in log lines (if
4803 the queue_time_overall log selector is set), a one-byte patch to exigrep
4804 was needed to allow it to recognize "Completed" as not the last thing in
4807 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4808 patch that reportedly fixes this has been added. I am not expert enough to
4809 create a test for it. This is what the patch creator wrote:
4811 "I found a little strange behaviour of ldap code when working with
4812 Windows 2003 AD Domain, where users was placed in more than one
4813 Organization Units. When I tried to give exim partial DN, the exit code
4814 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4815 But simultaneously result of request was absolutely normal ldap result,
4816 so I produce this patch..."
4818 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4819 REFERENCE, so I have modified the code to exclude the patch when that macro
4822 55. Some experimental protocols are using DNS PTR records for new purposes. The
4823 keys for these records are domain names, not reversed IP addresses. The
4824 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4825 leaves it alone. Component reversal etc. now happens only for IP addresses.
4828 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4830 57. Double the size of the debug message buffer (to 2048) so that more of very
4831 long debug lines gets shown.
4833 58. The exicyclog utility now does better if the number of log files to keep
4834 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4836 59. Two changes related to the smtp_active_hostname option:
4838 (1) $smtp_active_hostname is now available as a variable.
4839 (2) The default for smtp_banner uses $smtp_active_hostname instead
4840 of $primary_hostname.
4842 60. The host_aton() function is supposed to be passed a string that is known
4843 to be a valid IP address. However, in the case of IPv6 addresses, it was
4844 not checking this. This is a hostage to fortune. Exim now panics and dies
4845 if the condition is not met. A case was found where this could be provoked
4846 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4847 components; fortuitously, this particular loophole had already been fixed
4848 by change 4.50/55 above.
4850 If there are any other similar loopholes, the new check in host_aton()
4851 itself should stop them being exploited. The report I received stated that
4852 data on the command line could provoke the exploit when Exim was running as
4853 exim, but did not say which command line option was involved. All I could
4854 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4855 running as the user.
4858 61. There was a buffer overflow vulnerability in the SPA authentication code
4859 (which came originally from the Samba project). I have added a test to the
4860 spa_base64_to_bits() function which I hope fixes it.
4863 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4864 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4866 63. The daemon start-up calls getloadavg() while still root for those OS that
4867 need the first call to be done as root, but it missed one case: when
4868 deliver_queue_load_max is set with deliver_drop_privilege. This is
4869 necessary for the benefit of the queue runner, because there is no re-exec
4870 when deliver_drop_privilege is set.
4872 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4873 This has been fixed.
4875 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4876 was in use, was not putting the data itself into the right store pool;
4877 consequently, it could be overwritten for a subsequent message in the same
4878 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4881 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4883 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4884 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4885 the function were treating the return as a boolean value, which happened to
4886 work because 0=false and not-0=true, but is not correct code.
4888 68. The host_aton() function was not handling scoped IPv6 addresses (those
4889 with, for example, "%eth0" on the end) correctly.
4891 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4892 negated items (that is, ~something) in unsigned ints. Some compilers
4893 apparently mutter when there is no cast.
4895 70. If an address verification called from an ACL failed, and did not produce a
4896 user-specific message (i.e. there was only a "system" message), nothing was
4897 put in $acl_verify_message. In this situation, it now puts the system
4900 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4901 session; change 4.31/43 added the unwanted input to the log line - except
4902 that it did not do this in the start of session case. It now does.
4904 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4905 This is wrong and can cause the other end to generate a synchronization
4906 error if it is another Exim or anything else that does the synchronization
4907 check. A QUIT command is no longer sent after a timeout.
4909 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4910 during host lookups.
4912 74. The defer_ok option of callout verification was not working if it was used
4913 when verifying addresses in header lines, that is, for this case:
4915 verify = header_sender/callout=defer_ok
4917 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4918 those file descriptors could be used for SMTP connections. If anything
4919 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4920 could be sent to the SMTP client, causing chaos. The daemon now opens
4921 stdin, stdout, and stderr to /dev/null when it puts itself into the
4924 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4925 log by default. The user can override this with suitable Perl magic.
4927 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4928 the log message when discard triggers, was not working for the DATA ACL or
4929 for the non-SMTP ACL.
4931 78. Error message wording change in sieve.c.
4933 79. If smtp_accept_max_per_host was set, the number of connections could be
4934 restricted to fewer than expected, because the daemon was trying to set up
4935 a new connection before checking whether the processes handling previous
4936 connections had finished. The check for completed processes is now done
4937 earlier. On busy systems, this bug wouldn't be noticed because something
4938 else would have woken the daemon, and it would have reaped the completed
4941 80. If a message was submitted locally by a user whose login name contained one
4942 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4943 It caused a spool format error. I have fixed the spool reading code. A
4944 related problem was that the "from" clause in the Received: line became
4945 illegal because of the space(s). It is now covered by ${quote_local_part.
4947 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4950 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4952 83. Patch from Sieve maintainer to fix unterminated string problem in
4953 "vacation" handling.
4955 84. Some minor changes to the Linux configuration files to help with other
4956 OS variants using glibc.
4958 85. One more patch for Sieve to update vacation handling to latest spec.
4961 ----------------------------------------------------
4962 See the note above about the 4.44 and 4.50 releases.
4963 ----------------------------------------------------
4969 1. Change 4.43/35 introduced a bug that caused file counts to be
4970 incorrectly computed when quota_filecount was set in an appendfile
4973 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4974 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4977 3. Give more explanation in the error message when the command for a transport
4978 filter fails to execute.
4980 4. There are several places where Exim runs a non-Exim command in a
4981 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4982 was being done only for the command run by the queryprogram router. It is
4983 now done for all such subprocesses. The other cases are: ${run, transport
4984 filters, and the commands run by the lmtp and pipe transports.
4986 5. Some older OS have a limit of 256 on the maximum number of file
4987 descriptors. Exim was using setrlimit() to set 1000 as a large value
4988 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4989 systems. I've change it so that if it can't get 1000, it tries for 256.
4991 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4992 was an oversight, and furthermore, ever since the addition of extra
4993 controls (e.g. 4.43/32), the checks on when to allow different forms of
4994 "control" were broken. There should now be diagnostics for all cases when a
4995 control that does not make sense is encountered.
4997 7. $recipients is now available in the predata ACL (oversight).
4999 8. Tidy the search cache before the fork to do a delivery from a message
5000 received from the command line. Otherwise the child will trigger a lookup
5001 failure and thereby defer the delivery if it tries to use (for example) a
5002 cached ldap connection that the parent has called unbind on.
5004 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5005 of $address_data from the recipient verification was clobbered by the
5006 sender verification.
5008 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5009 was its contents. (It was OK if the option was not defined at all.)
5011 11. A "Completed" log line is now written for messages that are removed from
5012 the spool by the -Mrm option.
5014 12. $host_address is now set to the target address during the checking of
5015 ignore_target_hosts.
5017 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5018 being passed; this would have caused $sender_host_name to have been used if
5019 matching the list had actually called for a host name (not very likely,
5020 since this list is usually IP addresses). A host name is now passed as
5023 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5024 code that specifies a non-restarting handler (typically sigaction() in
5025 modern systems) in an attempt to fix a rare and obscure crash bug.
5027 15. Narrowed the window for a race in the daemon that could cause it to ignore
5028 SIGCHLD signals. This is not a major problem, because they are used only to
5029 wake it up if nothing else does.
5031 16. A malformed maildirsize file could cause Exim to calculate negative values
5032 for the mailbox size or file count. Odd effects could occur as a result.
5033 The maildirsize information is now recalculated if the size or filecount
5036 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5037 support for a long time. Removed HAVE_SYS_VFS_H.
5039 18. Updated exipick to current release from John Jetmore.
5041 19. Allow an empty sender to be matched against a lookup in an address list.
5042 Previously the only cases considered were a regular expression, or an
5045 20. Exim went into a mad DNS lookup loop when doing a callout where the
5046 host was specified on the transport, if the DNS lookup yielded more than
5049 21. The RFC2047 encoding function was originally intended for short strings
5050 such as real names; it was not keeping to the 75-character limit for
5051 encoded words that the RFC imposes. It now respects the limit, and
5052 generates multiple encoded words if necessary. To be on the safe side, I
5053 have increased the buffer size for the ${rfc2047: expansion operator from
5056 22. Failure to deliver a bounce message always caused it to be frozen, even if
5057 there was an errors_to setting on the router. The errors_to setting is now
5060 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5061 canonical form (fully expanded) before being placed in
5062 $sender_host_address.
5064 24. Updated eximstats to version 1.33
5066 25. Include certificate and key file names in error message when GnuTLS fails
5067 to set them up, because the GnuTLS error message doesn't include the name
5068 of the failing file when there is a problem reading it.
5070 26. Expand error message when OpenSSL has problems setting up cert/key files.
5073 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5074 (this can affect the format of dates).
5076 28. exim_tidydb, when checking for the continued existence of a message for
5077 which it has found a message-specific retry record, was not finding
5078 messages that were in split spool directories. Consequently, it was
5079 deleting retry records that should have stayed in existence.
5081 29. eximstats updated to version 1.35
5082 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5083 1.35 - bugfix such that pie charts by volume are generated correctly
5085 30. The SPA authentication driver was not abandoning authentication and moving
5086 on to the next authenticator when an expansion was forced to fail,
5087 contradicting the general specification for all authenticators. Instead it
5088 was generating a temporary error. It now behaves as specified.
5090 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5091 (the order specifies the preference for clients). The order is now AES256,
5092 AES128, 3DES, ARCFOUR128.
5094 31. Small patch to Sieve code - explicitly set From: when generating an
5097 32. Exim crashed if a remote delivery caused a very long error message to be
5098 recorded - for instance if somebody sent an entire SpamAssassin report back
5099 as a large number of 550 error lines. This bug was coincidentally fixed by
5100 increasing the size of one of Exim's internal buffers (big_buffer) that
5101 happened as part of the Exiscan merge. However, to be on the safe side, I
5102 have made the code more robust (and fixed the comments that describe what
5105 33. Some experimental protocols are using DNS PTR records for new purposes. The
5106 keys for these records are domain names, not reversed IP addresses. The
5107 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5108 leaves it alone. Component reversal etc. now happens only for IP addresses.
5111 34. The host_aton() function is supposed to be passed a string that is known
5112 to be a valid IP address. However, in the case of IPv6 addresses, it was
5113 not checking this. This is a hostage to fortune. Exim now panics and dies
5114 if the condition is not met. A case was found where this could be provoked
5115 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5116 components; fortuitously, this particular loophole had already been fixed
5117 by change 4.50/55 or 4.44/33 above.
5119 If there are any other similar loopholes, the new check in host_aton()
5120 itself should stop them being exploited. The report I received stated that
5121 data on the command line could provoke the exploit when Exim was running as
5122 exim, but did not say which command line option was involved. All I could
5123 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5124 running as the user.
5127 35. There was a buffer overflow vulnerability in the SPA authentication code
5128 (which came originally from the Samba project). I have added a test to the
5129 spa_base64_to_bits() function which I hope fixes it.
5132 36. The daemon start-up calls getloadavg() while still root for those OS that
5133 need the first call to be done as root, but it missed one case: when
5134 deliver_queue_load_max is set with deliver_drop_privilege. This is
5135 necessary for the benefit of the queue runner, because there is no re-exec
5136 when deliver_drop_privilege is set.
5138 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5139 was in use, was not putting the data itself into the right store pool;
5140 consequently, it could be overwritten for a subsequent message in the same
5141 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5144 38. Sometimes the final signoff response after QUIT could fail to get
5145 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5146 < 0 before doing a fflush(). This bug looks as though it goes back to the
5147 introduction of TLS in release 3.20, but "sometimes" must have been rare
5148 because the tests only now provoked it.
5154 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5155 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5156 Now it returns an integer. A number of calls were still expecting a T/F
5157 return. Fortuitously, in all cases, the tests worked in OK situations,
5158 which is the norm. However, things would have gone wrong on any write
5159 failures on the smtp file descriptor. This function is used when sending
5160 messages over SMTP and also when doing verify callouts.
5162 2. When Exim is called to do synchronous delivery of a locally submitted
5163 message (the -odf or -odi options), it no longer closes stderr before doing
5166 3. Implemented the mua_wrapper option.
5168 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5170 5. Implemented the functions header_remove(), header_testname(),
5171 header_add_at_position(), and receive_remove_recipient(), and exported them
5174 6. If an ACL "warn" statement specified the addition of headers, Exim already
5175 inserted X-ACL-Warn: at the start if there was no header name. However, it
5176 was not making this test for the second and subsequent header lines if
5177 there were newlines in the string. This meant that an invalid header could
5178 be inserted if Exim was badly configured.
5180 7. Allow an ACL "warn" statement to add header lines at the start or after all
5181 the Received: headers, as well as at the end.
5183 8. Added the rcpt_4xx retry error code.
5185 9. Added postmaster_mailfrom=xxx to callout verification option.
5187 10. Added mailfrom=xxxx to the callout verification option, for verify=
5190 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5191 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5193 12. Inserted some casts to stop certain compilers warning when using pointer
5194 differences as field lengths or precisions in printf-type calls (mostly
5195 affecting debugging statements).
5197 13. Added optional readline() support for -be (dynamically loaded).
5199 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5200 same clock tick as a message's arrival, so that its received time was the
5201 same as the "first fail" time on the retry record, and that message
5202 remained on the queue past the ultimate address timeout, every queue runner
5203 would try a delivery (because it was past the ultimate address timeout) but
5204 after another failure, the ultimate address timeout, which should have then
5205 bounced the address, did not kick in. This was a "< instead of <=" error;
5206 in most cases the first failure would have been in the next clock tick
5207 after the received time, and all would be well.
5209 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5210 being recognized when the domain list was tested by the match_domain
5211 condition in an expansion string.
5213 16. Added the ${str2b64: operator.
5215 17. Exim was always calling setrlimit() to set a large limit for the number of
5216 processes, without checking whether the existing limit was already
5217 adequate. (It did check for the limit on file descriptors.) Furthermore,
5218 errors from getrlimit() and setrlimit() were being ignored. Now they are
5219 logged to the main and panic logs, but Exim does carry on, to try to do its
5220 job under whatever limits there are.
5222 18. Imported PCRE 5.0.
5224 19. Trivial typo in log message " temporarily refused connection" (the leading
5227 20. If the log selector return_path_on_delivery was set and an address was
5228 redirected to /dev/null, the delivery process crashed because it assumed
5229 that a return path would always be set for a "successful" delivery. In this
5230 case, the whole delivery is bypassed as an optimization, and therefore no
5233 21. Internal re-arrangement: the function for sending a challenge and reading
5234 a response while authentication was assuming a zero-terminated challenge
5235 string. It's now changed to take a pointer and a length, to allow for
5236 binary data in such strings.
5238 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5240 23. Exim was not respecting finduser_retries when seeking the login of the
5241 uid under which it was called; it was always trying 10 times. (The default
5242 setting of finduser_retries is zero.) Also, it was sleeping after the final
5243 failure, which is pointless.
5245 24. Implemented tls_on_connect_ports.
5247 25. Implemented acl_smtp_predata.
5249 26. If the domain in control=submission is set empty, Exim assumes that the
5250 authenticated id is a complete email address when it generates From: or
5251 Sender: header lines.
5253 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5254 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5255 chown and chgrp in /bin and hostname in /usr/bin.
5257 28. Exim was keeping the "process log" file open after each use, just as it
5258 does for the main log. This opens the possibility of it remaining open for
5259 long periods when the USR1 signal hits a daemon. Occasional processlog
5260 errors were reported, that could have been caused by this. Anyway, it seems
5261 much more sensible not to leave this file open at all, so that is what now
5264 29. The long-running daemon process does not normally write to the log once it
5265 has entered its main loop, and it closes the log before doing so. This is
5266 so that log files can straightforwardly be renamed and moved. However,
5267 there are a couple of unusual error situations where the daemon does write
5268 log entries, and I had neglected to close the log afterwards.
5270 30. The text of an SMTP error response that was received during a remote
5271 delivery was being truncated at 512 bytes. This is too short for some of
5272 the long messages that one sometimes sees. I've increased the limit to
5275 31. It is now possible to make retry rules that apply only when a message has a
5276 specific sender, in particular, an empty sender.
5278 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5279 it possible to be selective about when SMTP synchronization is enforced.
5281 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5283 32. Implemented hosts_connection_nolog.
5285 33. Added an ACL for QUIT.
5287 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5290 35. Added mailbox_size and mailbox_filecount to appendfile.
5292 36. Added control = no_multiline_responses to ACLs.
5294 37. There was a bug in the logic of the code that waits for the clock to tick
5295 in the case where the clock went backwards by a substantial amount such
5296 that the microsecond fraction of "now" was more than the microsecond
5297 fraction of "then" (but the whole seconds number was less).
5299 38. Added support for the libradius Radius client library this is found on
5300 FreeBSD (previously only the radiusclient library was supported).
5306 1. When certain lookups returned multiple values in the form name=value, the
5307 quoting of the values was not always being done properly. Specifically:
5308 (a) If the value started with a double quote, but contained no whitespace,
5310 (b) If the value contained whitespace other than a space character (i.e.
5311 tabs or newlines or carriage returns) it was not quoted.
5312 This fix has been applied to the mysql and pgsql lookups by writing a
5313 separate quoting function and calling it from the lookup code. The fix
5314 should probably also be applied to nisplus, ibase and oracle lookups, but
5315 since I cannot test any of those, I have not disturbed their existing code.
5317 2. A hit in the callout cache for a specific address caused a log line with no
5318 reason for rejecting RCPT. Now it says "Previous (cached) callout
5319 verification failure".
5321 3. There was an off-by-one bug in the queryprogram router. An over-long
5322 return line was truncated at 256 instead of 255 characters, thereby
5323 overflowing its buffer with the terminating zero. As well as fixing this, I
5324 have increased the buffer size to 1024 (and made a note to document this).
5326 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5327 when Exim is waiting for an SMTP response from a remote server, Exim
5328 restarts its select() call on the socket, thereby resetting its timeout.
5329 This is not a problem when such interrupts are rare. Somebody set up a cron
5330 job to run exiwhat every 2 minutes, which is less than the normal select()
5331 timeout (5 or 10 minutes). This meant that the select() timeout never
5332 kicked in because it was always reset. I have fixed this by comparing the
5333 time when an interrupt arrives with the time at the start of the first call
5334 to select(). If more time than the timeout has elapsed, the interrupt is
5335 treated as a timeout.
5337 5. Some internal re-factoring in preparation for the addition of Sieve
5338 extensions (by MH). In particular, the "personal" test is moved to a
5339 separate function, and given an option for scanning Cc: and Bcc: (which is
5340 not set for Exim filters).
5342 6. When Exim created an email address using the login of the caller as the
5343 local part (e.g. when creating a From: or Sender: header line), it was not
5344 quoting the local part when it contained special characters such as @.
5346 7. Installed new OpenBSD configuration files.
5348 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5349 try to make them clearer.
5351 9. Callout options, other than the timeout value, were being ignored when
5352 verifying sender addresses in header lines. For example, when using
5354 verify = header_sender/callout=no_cache
5356 the cache was (incorrectly) being used.
5358 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5359 only the Cygwin environment.
5361 11. When return_path_on_delivery was set as a log selector, if different remote
5362 addresses in the same message used different return paths and parallel
5363 remote delivery occurred, the wrong values would sometimes be logged.
5364 (Whenever a remote delivery process finished, the return path value from
5365 the most recently started remote delivery process was logged.)
5367 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5368 lines when AUTH and/or TLS are in use. This is the "received protocol"
5369 field. Exim used to use "asmtp" for authenticated SMTP, without any
5370 indication (in the protocol name) for TLS use. Now it follows the RFC and
5371 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5372 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5373 names appear in log lines as well as in Received: header lines.
5375 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5376 extensions, and comparison tests, and to fix some bugs.
5378 14. Changes to the "personal" filter test:
5380 (1) The test was buggy in that it was just doing the equivalent of
5381 "contains" tests on header lines. For example, if a user's address was
5382 anne@some.where, the "personal" test would incorrectly be true for
5384 To: susanne@some.where
5386 This test is now done by extracting each address from the header in turn,
5387 and checking the entire address. Other tests that are part of "personal"
5388 are now done using regular expressions (for example, to check local parts
5389 of addresses in From: header lines).
5391 (2) The list of non-personal local parts in From: addresses has been
5392 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5393 taken from the Sieve specification recommendations.
5395 (3) If the message contains any header line starting with "List-" it is
5396 treated as non-personal.
5398 (4) The test for "circular" in the Subject: header line has been removed
5399 because it now seems ill-conceived.
5401 15. Minor typos in src/EDITME comments corrected.
5403 16. Installed latest exipick from John Jetmore.
5405 17. If headers_add on a router specified a text string that was too long for
5406 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5407 of string_sprintf() is now avoided.
5409 18. $message_body_size was not set (it was always zero) when running the DATA
5410 ACL and the local_scan() function.
5412 19. For the "mail" command in an Exim filter, no default was being set for
5413 the once_repeat time, causing a random time value to be used if "once" was
5414 specified. (If the value happened to be <= 0, no repeat happened.) The
5415 default is now 0s, meaning "never repeat". The "vacation" command was OK
5416 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5417 (I found it when inspecting the code).
5419 20. There is now an overall timeout for performing a callout verification. It
5420 defaults to 4 times the callout timeout, which applies to individual SMTP
5421 commands during the callout. The overall timeout applies when there is more
5422 than one host that can be tried. The timeout is checked before trying the
5423 next host. This prevents very long delays if there are a large number of
5424 hosts and all are timing out (e.g. when the network connections are timing
5425 out). The value of the overall timeout can be changed by specifying an
5426 additional sub-option for "callout", called "maxwait". For example:
5428 verify = sender/callout=5s,maxwait=20s
5430 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5431 to the end before writing, but this should make it even safer).
5433 22. Exim was forgetting that it had advertised PIPELINING for the second and
5434 subsequent messages on an SMTP connection. It was also not resetting its
5435 memory on STARTTLS and an internal HELO.
5437 23. When Exim logs an SMTP synchronization error within a session, it now
5438 records whether PIPELINING has been advertised or not.
5440 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5441 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5442 rather than long int.
5444 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5446 26. Added the never_mail option to autoreply.
5452 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5453 crash if the getsockname() call failed; this can happen if a connection is
5454 closed very soon after it is established. The problem was simply in the
5455 order in which certain operations were done, causing Exim to try to write
5456 to the SMTP stream before it had set up the file descriptor. The bug has
5457 been fixed by making things happen in the correct order.
5463 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5464 before the connection was closed, thus losing the rejection response.
5466 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5467 some early Solaris releases, but causes trouble in current releases where
5468 socklen_t is defined.
5470 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5473 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5476 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5477 that releasing the top of it at the end releases what was used for sub-
5478 expansions (unless the block got too big). However, discard this block if
5479 the first thing is a variable or header, so that we can use its block when
5480 it is dynamic (useful for very large $message_headers, for example).
5482 6. Lookups now cache *every* query, not just the most recent. A new, separate
5483 store pool is used for this. It can be recovered when all lookup caches are
5484 flushed. Lookups now release memory at the end of their result strings.
5485 This has involved some general refactoring of the lookup sources.
5487 7. Some code has been added to the store_xxx() functions to reduce the amount
5488 of flapping under certain conditions.
5490 8. log_incoming_interface used to affect only the <= reception log lines. Now
5491 it causes the local interface and port to be added to several more SMTP log
5492 lines, for example "SMTP connection from", and rejection lines.
5494 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5496 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5498 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5499 could be overwritten at the end of the current message (or the start of a
5500 new message if it was set in a HELO ACL). The value is now preserved for
5501 the duration of the SMTP connection.
5503 12. If a transport had a headers_rewrite setting, and a matching header line
5504 contained an unqualified address, that address was qualified, even if it
5505 did not match any rewriting rules. The underlying bug was that the values
5506 of the flags that permit the existence of unqualified sender and recipient
5507 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5508 non-local messages, and by -bnq for local messages) were not being
5509 preserved with the message after it was received.
5511 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5512 "next input=" as part of the text comprising the host identity instead of
5513 the correct text. The code was using the same buffer for two different
5514 strings. However, depending on which order the printing function evaluated
5515 its arguments, the bug did not always show up. Under Linux, for example, my
5516 test suite worked just fine.
5518 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5519 doesn't work with some older versions of Perl. It has been changed to "my",
5520 which in any case is probably the better facility to use.
5522 15. A really picky compiler found some instances of statements for creating
5523 error messages that either had too many or two few arguments for the format
5526 16. The size of the buffer for calls to the DNS resolver has been increased
5527 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5528 for addresses that have a lot of PTR records. This alleviates a problem; it
5529 does not fully solve it.
5531 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5532 buffer now truncates the list and logs the incident, which is the same
5533 action as happens when Exim is looking up a host name and its aliases.
5534 Previously in this situation something unpredictable would happen;
5535 sometimes it was "internal error: store_reset failed".
5537 18. If a server dropped the connection unexpectedly when an Exim client was
5538 using GnuTLS and trying to read a response, the client delivery process
5539 crashed while trying to generate an error log message.
5541 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5542 string, for example:
5544 warn message = H1: something\nH2: something
5546 the text was added as a single header line from Exim's point of view
5547 though it ended up OK in the delivered message. However, searching for the
5548 second and subsequent header lines using $h_h2: did not work. This has been
5549 fixed. Similarly, if a system filter added multiple headers in this way,
5550 the routers could not see them.
5552 20. Expanded the error message when iplsearch is called with an invalid key to
5553 suggest using net-iplsearch in a host list.
5555 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5556 ACLs are no longer actually imposed (and a message to that effect is
5559 22. If a "gecos" field in a passwd entry contained escaped characters, in
5560 particular, if it contained a \" sequence, Exim got it wrong when building
5561 a From: or a Sender: header from that name. A second bug also caused
5562 incorrect handling when an unquoted " was present following a character
5563 that needed quoting.
5565 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5566 was not being matched caselessly.
5568 24. Arranged for all hyphens in the exim.8 source to be escaped with
5571 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5572 properly. Recipient callout cache records were still being keyed to include
5573 the sender, even when use_sender was set false. This led to far more
5574 callouts that were necessary. The sender is no longer included in the key
5575 when use_sender is false.
5577 26. Added "control = submission" modifier to ACLs.
5579 27. Added the ${base62d: operator to decode base 62 numbers.
5581 28. dnsdb lookups can now access SRV records.
5583 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5584 the configuration file.
5586 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5587 (-v) mode. This makes the output for a verbose queue run more intelligible.
5589 31. Added a use_postmaster feature to recipient callouts.
5591 32. Added the $body_zerocount variable, containing the number of binary zero
5592 bytes in the message body.
5594 33. The time of last modification of the "new" subdirectory is now used as the
5595 "mailbox time last read" when there is a quota error for a maildir
5598 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5600 35. Added +ignore_unknown as a special item in host lists.
5602 36. Code for decoding IPv6 addresses in host lists is now included, even if
5603 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5604 address was recognized as an IP address, but was then not correctly decoded
5605 into binary, causing unexpected and incorrect effects when compared with
5612 1. Very minor rewording of debugging text in manualroute to say "list of
5613 hosts" instead of "hostlist".
5615 2. If verify=header_syntax was set, and a header line with an unqualified
5616 address (no domain) and a large number of spaces between the end of the
5617 name and the colon was received, the reception process suffered a buffer
5618 overflow, and (when I tested it) crashed. This was caused by some obsolete
5619 code that should have been removed. The fix is to remove it!
5621 3. When running in the test harness, delay a bit after writing a bounce
5622 message to get a bit more predictability in the log output.
5624 4. Added a call to search_tidyup() just before forking a reception process. In
5625 theory, someone could use a lookup in the expansion of smtp_accept_max_
5626 per_host which, without the tidyup, could leave open a database connection.
5628 5. Added the variables $recipient_data and $sender_data which get set from a
5629 lookup success in an ACL "recipients" or "senders" condition, or a router
5630 "senders" option, similar to $domain_data and $local_part_data.
5632 6. Moved the writing of debug_print from before to after the "senders" test
5635 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5636 problems for message scanning, either using a data ACL, or using
5637 local_scan() because the Received: header was not generated till after they
5638 were called (in order to set the time as the time of reception completion).
5639 I have revised the way this works. The header is now generated after the
5640 body is received, but before the ACL or local_scan() are called. After they
5641 are run, the timestamp in the header is updated.
5647 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5648 before starting a queue runner without re-exec. This happened only when
5649 deliver_drop_privilege was set or when the Exim user was set to root. The
5650 effect of the bug was that timeouts during subsequent deliveries caused
5651 crashes instead of being properly handled. The handler is now left at its
5652 default (and expected) setting.
5654 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5655 message, again when deliver_drop_privilege is set or Exim is run as root.
5656 The bug described in (1) was not present in this case, but the tidying up
5657 of the other signals was missing. I have made the two cases consistent.
5659 3. The ignore_target_hosts setting on a manualroute router was being ignored
5660 for hosts that were looked up using the /MX notation.
5662 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5665 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5666 operated on the sender address. After changing the $sender_address to <>
5667 for the sender address verify, Exim was re-instated it as the original
5668 (before rewriting) address, but remembering that it had rewritten it, so it
5669 wasn't rewriting it again. This bug also had the effect of breaking the
5670 sender address verification caching when the sender address was rewritten.
5672 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5673 This has been changed so that if the ip literal address matches
5674 ignore_target_hosts, the router declines.
5676 7. Added expansion conditions match_domain, match_address, and match_local_
5677 part (NOT match_host).
5679 8. The placeholder for the Received: header didn't have a length field set.
5681 9. Added code to Exim itself and to exim_lock to test for a specific race
5682 condition that could lead to file corruption when using MBX delivery. The
5683 issue is with the lockfile that is created in /tmp. If this file is removed
5684 after a process has opened it but before that process has acquired a lock,
5685 there is the potential for a second process to recreate the file and also
5686 acquire a lock. This could lead to two Exim processes writing to the file
5687 at the same time. The added code performs the same test as UW imapd; it
5688 checks after acquiring the lock that its file descriptor still refers to
5689 the same named file.
5691 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5692 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5695 11. Added the smtp_active_hostname option. If used, this will typically be made
5696 to depend on the incoming interface address. Because $interface_address is
5697 not set up until the daemon has forked a reception process, error responses
5698 that can happen earlier (such as "too many connections") no longer contain
5701 12. If an expansion in a condition on a "warn" statement fails because a lookup
5702 defers, the "warn" statement is abandoned, and the next ACL statement is
5703 processed. Previously this caused the whole ACL to be aborted.
5705 13. Added the iplsearch lookup type.
5707 14. Added ident_timeout as a log selector.
5709 15. Added tls_certificate_verified as a log selector.
5711 16. Added a global option tls_require_ciphers (compare the smtp transport
5712 option of the same name). This controls incoming TLS connections.
5714 17. I finally figured out how to make tls_require_ciphers do a similar thing
5715 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5716 before starting the TLS session.
5718 18. Tabs are now shown as \t in -bP output.
5720 19. If the log selector return_path_on_delivery was set, Exim crashed when
5721 bouncing a message because it had too many Received: header lines.
5723 20. If two routers both had headers_remove settings, and the first one included
5724 a superfluous trailing colon, the final name in the first list and the
5725 first name in the second list were incorrectly joined into one item (with a
5726 colon in the middle).
5732 1. Added -C and -D options to the exinext utility, mainly to make it easier
5733 to include in the automated testing, but these could be helpful when
5734 multiple configurations are in use.
5736 2. The exinext utility was not formatting the output nicely when there was
5737 an alternate port involved in the retry record key, nor when there was a
5738 message id as well (for retries that were specific to a specific message
5739 and a specific host). It was also confused by IPv6 addresses, because of
5740 the additional colons they contain. I have fixed the IPv4 problem, and
5741 patched it up to do a reasonable job for IPv6.
5743 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5744 delivery, the log line now contains "pipelined" if PIPELINING was used.
5746 4. An SMTP transport process used to panic and die if the bind() call to set
5747 an explicit outgoing interface failed. This has been changed; it is now
5748 treated in the same way as a connect() failure.
5750 5. A reference to $sender_host_name in the part of a conditional expansion
5751 that was being skipped was still causing a DNS lookup. This no longer
5754 6. The def: expansion condition was not recognizing references to header lines
5755 that used bh_ and bheader_.
5757 7. Added the _cache feature to named lists.
5759 8. The code for checking quota_filecount in the appendfile transport was
5760 allowing one more file than it should have been.
5762 9. For compatibility with Sendmail, the command line option
5770 and sets the incoming protocol and host name (for trusted callers). The
5771 host name and its colon can be omitted when only the protocol is to be set.
5772 Note the Exim already has two private options, -pd and -ps, that refer to
5773 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5774 "s", but I don't think that's a major issue.
5776 10. A number of refactoring changes to the code, none of which should affect
5779 (a) The number of logging options was getting close to filling up the
5780 32-bit word that was used as a bit map. I have split them into two classes:
5781 those that are passed in the argument to log_write(), and those that are
5782 only ever tested independently outside of that function. These are now in
5783 separate 32-bit words, so there is plenty of room for expansion again.
5784 There is no change in the user interface or the logging behaviour.
5786 (b) When building, for example, log lines, the code previously used a
5787 macro that called string_cat() twice, in order to add two strings. This is
5788 not really sufficiently general. Furthermore, there was one instance where
5789 it was actually wrong because one of the argument was used twice, and in
5790 one call a function was used. (As it happened, calling the function twice
5791 did not affect the overall behaviour.) The macro has been replaced by a
5792 function that can join an arbitrary number of extra strings onto a growing
5795 (c) The code for expansion conditions now uses a table and a binary chop
5796 instead of a serial search (which was left over from when there were very
5797 few conditions). Also, it now recognizes conditions like "pam" even when
5798 the relevant support is not compiled in: a suitably worded error message is
5799 given if an attempt is made to use such a condition.
5801 11. Added ${time_interval:xxxxx}.
5803 12. A bug was causing one of the ddress fields not to be passed back correctly
5804 from remote delivery subprocesses. The field in question was not being
5805 subsequently used, so this caused to problems in practice.
5807 13. Added new log selectors queue_time and deliver_time.
5809 14. Might have fixed a bug in maildirsizefile handling that threw up
5810 "unexpected character" debug warnings, and recalculated the data
5811 unnecessarily. In any case, I expanded the warning message to give more
5814 15. Added the message "Restricted characters in address" to the statements in
5815 the default ACL that block characters like @ and % in local parts.
5817 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5818 Three changes have been made:
5820 (a) There was a serious bug; a negative response to MAIL caused the whole
5821 recipient domain to be cached as invalid, thereby blocking all messages
5822 to all local parts at the same domain, from all senders. This bug has
5823 been fixed. The domain is no longer cached after a negative response to
5824 MAIL if the sender used is not empty.
5826 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5829 (c) A new callout option, "use_sender" has been added for people who want
5830 the modified behaviour.
5836 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5839 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5840 indeed breaks things for older releases.
5842 3. Added additional logging to the case where there is a problem reading data
5843 from a filter that is running in a subprocess using a pipe, in order to
5844 try to track down a specific problem.
5846 4. Testing facility fudge: when running in the test harness and attempting
5847 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5848 getting "No route to host". Convert this to a timeout.
5850 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5853 6. Some OS don't have socklen_t but use size_t instead. This affects the
5854 fifth argument of getsockopt() amongst other things. This is now
5855 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5856 can be set for individual OS. I have set it for SunOS5, OSF1, and
5857 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5858 some earlier ones do not.
5860 7. Change 4.30/15 was not doing the test caselessly.
5862 8. The standard form for an IPv6 address literal was being rejected by address
5863 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5864 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5865 this, as well as the form without the "IPv6" on the front (but only when
5866 address literals are enabled, of course).
5868 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5870 10. Exim crashed if a message with an empty sender address specified by -f
5871 encountered a router with an errors_to setting. This could be provoked only
5872 by a command such as
5876 where an empty string was supplied; "<>" did not hit this bug.
5878 11. Installed PCRE release 4.5.
5880 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5881 remained set. It is now erased.
5883 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5884 times from message ids (which are base 36 rather than the normal 62).
5886 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5887 were being counted as actual protocol errors, and logged if the log
5888 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5889 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5890 and DATA following a set of rejected RCPTs do not count as protocol errors.
5891 In other words, Exim assumes they were pipelined, though this may not
5892 actually be the case. Of course, in all cases the client gets an
5893 appropriate error code.
5895 15. If a lookup fails in an ACL condition, a message about the failure may
5896 be available; it is used if testing the ACL cannot continue, because most
5897 such messages specify what the cause of the deferral is. However, some
5898 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5899 that caused an old message to be retained and used if a later statement
5900 caused a defer, replacing the real cause of the deferral.
5902 16. If an IP address had so many PTR records that the DNS lookup buffer
5903 was not large enough to hold them, Exim could crash while trying to process
5904 the truncated data. It now detects and logs this case.
5906 17. Further to 4.21/58, another change has been made: if (and only if) the
5907 first line of a message (the first header line) ends with CRLF, a bare LF
5908 in a subsequent header line has a space inserted after it, so as not to
5909 terminate the header.
5911 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5912 unnecessarily, used atoi() instead of strtol(), and didn't check the
5913 termination when getting file sizes from file names by regex.
5915 19. Completely re-implemented the support for maildirsize files, in the light
5916 of a number of problems with the previous contributed implementation
5917 (4.30/29). In particular:
5919 . If the quota is zero, the maildirsize file is maintained, but no quota is
5922 . If the maildir directory does not exist, it is created before any attempt
5923 to write a maildirsize file.
5925 . The quota value in the file is just a cache; if the quota is changed in
5926 the transport, the new value overrides.
5928 . A regular expression is available for excluding directories from the
5931 20. The autoreply transport checks the characters in options that define the
5932 message's headers; it allows continued headers, but it was checking with
5933 isspace() after an embedded newline instead of explicitly looking for a
5936 21. If all the "regular" hosts to which an address was routed had passed their
5937 expiry times, and had not reached their retry times, the address was
5938 bounced, even if fallback hosts were defined. Now Exim should go on to try
5941 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5942 equivalent code in the SMTP transport. Some hosts send humungous responses
5943 to HELO/EHLO, more than 1024 it seems.
5945 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5946 gives compiler warnings in some environments. I've now done it "properly",
5949 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5950 (because of problems with the built-in one) was declared to return uschar *
5951 instead of char *, causing compiler failure.
5953 25. Fixed a file descriptor leak when processing alias/forward files.
5955 26. Fixed a minor format string issue in dbfn.c.
5957 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5959 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5960 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5963 29. When a sender address is verified, it is cached, to save repeating the test
5964 when there is more than one recipient in a message. However, when the
5965 verification involves a callout, it is possible for different callout
5966 options to be set for different recipients. It is too complicated to keep
5967 track of this in the cache, so now Exim always runs a verification when a
5968 callout is required, relying on the callout cache for the optimization.
5969 The overhead is duplication of the address routing, but this should not be
5972 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5973 to be verified with callout=postmaster, and the main callout worked but the
5974 postmaster check failed, the verification correctly failed. However, if a
5975 subsequent RCPT command asked for sender verification *without* the
5976 postmaster check, incorrect caching caused this verification also to fail,
5979 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5980 it was not caching the DNS options (qualify_single, search_parents) that
5981 were used when the lookup failed. A subsequent lookup with different
5982 options therefore always gave the same answer, though there were cases
5983 where it should not have. (Example: a "domains = !$mx_any" option on a
5984 dnslookup router: the "domains" option is always processed without any
5985 widening, but the router might have qualify_single set.) Now Exim uses the
5986 cached value only when the same options are set.
5988 32. Added John Jetmore's "exipick" utility to the distribution.
5990 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5991 than a timeout (e.g. a certificate is required, and is not provided), an
5992 Exim server now closes the connection immediately. Previously it waited for
5993 the client to close - but if the client is SSL, it seems that they each
5994 wait for each other, leading to a delay before one of them times out.
5996 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5997 maintained 0.8.x compatibility because I don't think many are using it, and
5998 it is clearly obsolete.
6000 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6003 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6004 client certificate was expired. A simple patch fixes this, though I don't
6005 understand the full logic of why the verify callback is called multiple
6008 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6009 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6010 which causes problems with some clients (such as the Certicom SSL Plus
6011 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6012 disables the coutermeasure allowing Eudora to connect."
6014 38. Exim was not checking that a write() to a log file succeeded. This could
6015 lead to Bad Things if a log got too big, in particular if it hit a file
6016 size limit. Exim now panics and dies if it cannot write to a log file, just
6017 as it does if it cannot open a log file.
6019 39. Modified OS/Makefile-Linux so that it now contains
6021 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6023 The two -D definitions ensure that Exim is compiled with large file
6024 support, which makes it possible to handle log files that are bigger than
6027 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6028 instance) a domain was checked against a named list that involved a lookup,
6029 causing $domain_data to be set, then another domain was checked against the
6030 same list, then the first domain was re-checked, the value of $domain_data
6031 after the final check could be wrong. In particular, if the second check
6032 failed, it could be set empty. This bug probably also applied to
6035 41. The strip_trailing_dot option was not being applied to the address given
6036 with the -f command-line option.
6038 42. The code for reading a message's header from the spool was incrementing
6039 $received_count, but never initializing it. This meant that the value was
6040 incorrect (doubled) while delivering a message in the same process in which
6041 it was received. In the most common configuration of Exim, this never
6042 happens - a fresh exec is done - but it can happen when
6043 deliver_drop_privilege is set.
6045 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6046 it now includes up to 150 characters of the unexpected data in the log
6049 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6050 and building data strings. The size of both of these buffers was 10 000
6051 bytes - far larger than anybody would *ever* want, thought I. Needless to
6052 say, somebody hit the limit. I have increased the maximum line length to
6053 20 000 and the maximum data length of concatenated lines to 100 000. I have
6054 also fixed two bugs, because there was no checking on these buffers. Tsk,
6055 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6056 buffer is too small.
6058 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6059 lsearch lookups. Now it does.
6061 46. When parsing a route_list item in a manualroute router, a fixed-length
6062 buffer was used for the list of hosts. I made this 1024 bytes long,
6063 thinking that nobody would ever have a list of hosts that long. Wrong.
6064 Somebody had a whole pile of complicated expansion conditions, and the
6065 string was silently truncated, leading to an expansion error. It turns out
6066 that it is easier to change to an unlimited length (owing to other changes
6067 that have happened since this code was originally written) than to build
6068 structure for giving a limitation error. The length of the item that
6069 expands into the list of hosts is now unlimited.
6071 47. The lsearch lookup could not handle data where the length of text line was
6072 more than 4095 characters. Such lines were truncated, leading to shortened
6073 data being returned. It should now handle lines of any length.
6075 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6076 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6079 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6080 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6081 no longer happens in comment lines. A list of replacements is now placed
6082 at the head of all of the source files, except those whose only change is
6083 to replace PERL_COMMAND in the very first #! line.
6085 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6086 messages on the queue, with a bottom-up merge sort, using code contributed
6087 by Michael Haardt. This should make operations like -bp somewhat faster on
6088 large queues. It won't affect queue runners, except when queue_run_in_order
6091 51. Installed eximstats 1.31 in the distribution.
6093 52. Added support for SRV lookups to the dnslookup router.
6095 53. If an ACL referred to $message_body or $message_body_end, the value was not
6096 reset for any messages that followed in the same SMTP session.
6098 54. The store-handling optimization for building very long strings was not
6099 differentiating between the different store pools. I don't think this
6100 actually made any difference in practice, but I've tidied it.
6102 55. While running the routers to verify a sender address, $sender_address
6103 was still set to the sender address. This is wrong, because when routing to
6104 send a bounce to the sender, it would be empty. Therefore, I have changed
6105 it so that, while verifying a sender address, $sender_address is set to <>.
6106 (There is no change to what happens when verifying a recipient address.)
6108 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6109 target A or AAAA records (if not already returned) without resetting the
6110 qualify_single or search_parents options of the DNS resolver. These are
6111 inappropriate in this case because the targets of MX and SRV records must
6112 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6113 target that, when qualified, matched something in the local domain. These
6114 two options are now turned off when doing these lookups.
6116 57. It seems that at least some releases of Reiserfs (which does not have the
6117 concept of a fixed number of inodes) returns zero and not -1 for the
6118 number of available inodes. This interacted badly with check_spool_inodes,
6119 which assumed that -1 was the "no such thing" setting. What I have done is
6120 to check that the total number of inodes is greater than zero before doing
6121 the test of how many are available.
6123 58. When a "warn" ACL statement has a log_message modifier, the message is
6124 remembered, and not repeated. This is to avoid a lot of repetition when a
6125 message has many recipients that cause the same warning to be written.
6126 However, Exim was preserving the list of already written lines for an
6127 entire SMTP session, which doesn't seem right. The memory is now reset if a
6128 new message is started.
6130 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6131 debugging output unless log_rewrite was also set.
6133 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6134 of (int)(handle) when we know that handle contains (void *)(-1).
6136 61. The Exim daemon panic-logs an error return when it closes the incoming
6137 connection. However "connection reset by peer" seems to be common, and
6138 isn't really an error worthy of noting specially, so that particular error
6141 62. When Exim is trying to find all the local interfaces, it used to panic and
6142 die if the ioctl to get the interface flags failed. However, it seems that
6143 on at least one OS (Solaris 9) it is possible to have an interface that is
6144 included in the list of interfaces, but for which you get a failure error
6145 for this call. This happens when the interface is not "plumbed" into a
6146 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6147 failure of the "get flags" call assumes that the interface is down.
6149 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6150 makes life easier for people who are doing arithmetic on fields extracted
6151 from dates, where you often get leading zeros that should not be
6152 interpreted as octal.
6154 64. Added qualify_domain to the redirect router, to override the global
6157 65. If a pathologically long header line contained very many addresses (the
6158 report of this problem mentioned 10 000) and each of them was rewritten,
6159 Exim could use up a very large amount of memory. (It kept on making new
6160 copies of the header line as it rewrote, and never released the old ones.)
6161 At the expense of a bit more processing, the header rewriting function has
6162 been changed so that it no longer eats memory in this way.
6164 66. The generation of the Received: header has been moved from the time that a
6165 message starts to be received, to the time that it finishes. The timestamp
6166 in the Received: header should now be very close to that of the <= log
6167 line. There are two side-effects of this change:
6169 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6170 logged header lines no longer include the local Received: line, because
6171 it has not yet been created. The same applies to a copy of the message
6172 that is returned to a non-SMTP sender when a message is rejected.
6174 (b) When a filter file is tested using -bf, no additional Received: header
6175 is added to the test message. After some thought, I decided that this
6178 This change does not affect the value of $received_for. It is still set
6179 after address rewriting, but before local_scan() is called.
6181 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6183 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6184 gave an unhelpful panic error message, and a defer error. I have managed to
6185 change this behaviour so that it now rejects any supplied certificate,
6186 which seems right, as the list of acceptable certificates is empty.
6188 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6189 gave an unhelpful defer error. I have not managed to make this reject any
6190 supplied certificates, but the error message it gives is "no certificate
6191 supplied", which is not helpful.
6193 70. exigrep's output now also includes lines that are not associated with any
6194 message, but which match the given pattern. Implemented by a patch from
6195 Martin Sluka, which also tidied up the Perl a bit.
6197 71. Recipient callout verification, like sender verification, was using <> in
6198 the MAIL FROM command. This isn't really the right thing, since the actual
6199 sender may affect whether the remote host accepts the recipient or not. I
6200 have changed it to use the actual sender in the callout; this means that
6201 the cache record is now keyed on a recipient/sender pair, not just the
6202 recipient address. There doesn't seem to be a real danger of callout loops,
6203 since a callout by the remote host to check the sender would use <>.
6204 [SEE ABOVE: changed after hitting problems.]
6206 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6207 temporary errors. However, in the case of such a code being given after
6208 the end of a data transmission (i.e. after ".") Exim was failing to write
6209 a retry record for the message. (Yes, there was some broken host that was
6210 actually sending 8xx at this point.)
6212 73. An unknown lookup type in a host list could cause Exim to panic-die when
6213 the list was checked. (An example that provoked this was putting <; in the
6214 middle of a list instead of at the start.) If this happened during a DATA
6215 ACL check, a -D file could be left lying around. This kind of configuration
6216 error no longer causes Exim to die; instead it causes a defer error. The
6217 incident is still logged to the main and panic logs.
6219 74. Buglet left over from Exim 3 conversion. The message "too many messages
6220 in one connection" was written to the rejectlog but not the mainlog, except
6221 when address rewriting (yes!) was being logged.
6223 75. Added write_rejectlog option.
6225 76. When a system filter was run not as root (that is, when system_filter_user
6226 was set), the values of the $n variables were not being returned to the
6227 main process; thus, they were not subsequently available in the $sn
6230 77. Added +return_path_on_delivery log selector.
6232 78. A connection timeout was being treated differently from recipients deferred
6233 when testing hosts_max_try with a message that was older than the host's
6234 retry timeout. (The host should not be counted, thus allowing all hosts to
6235 be tried at least once before bouncing.) This may have been the cause of an
6236 occasionally reported bug whereby a message would remain on the queue
6237 longer than the retry timeout, but would be bounced if a delivery was
6238 forced. I say "may" because I never totally pinned down the problem;
6239 setting up timeout/retry tests is difficult. See also the next item.
6241 79. The ultimate address timeout was not being applied to errors that involved
6242 a combination of host plus message (for example, a timeout on a MAIL
6243 command). When an address resolved to a number of possible hosts, and they
6244 were not all tried for each delivery (e.g. because of hosts_max_try), a
6245 message could remain on the queue longer than the retry timeout.
6247 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6250 81. Fixed an obscure SMTP outgoing bug which required at least the following
6251 conditions: (a) there was another message waiting for the same server;
6252 (b) the server returned 5xx to all RCPT commands in the first message so
6253 that the message was not completed; (c) the server dropped the connection
6254 or gave a negative response to the RSET that Exim sends to abort the
6255 transaction. The observed case was a dropped connection after DATA that had
6256 been sent in pipelining mode. That is, the server had advertised PIPELINING
6257 but was not implementing it correctly. The effect of the bug was incorrect
6258 behaviour, such as trying another host, and this could lead to a crash.
6264 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6265 and daemon.c were passed as pointers to ints; they should have been
6266 pointers to socklen_t variables (which are typically unsigned ints).
6268 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6271 3. Fixed a really odd bug that affected only the testing scheme; patching a
6272 certain fixed string in the binary changed the value of another string that
6273 happened to be identical to the end of the original first string.
6275 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6276 name", it returns that address as the IP address. On some operating
6277 systems (e.g. Solaris), it also passes back the IP address string as the
6278 "host name". However, on others (e.g. Linux), it passes back an empty
6279 string. Exim wasn't checking for this, and was changing the host name to an
6280 empty string, assuming it had been canonicalized.
6282 5. Although rare, it is permitted to have more than one PTR record for a given
6283 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6284 all the names associated with an address, because they do in Solaris.
6285 However, it seems that they do not in Linux for data that comes from the
6286 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6287 I found this out when I moved to a new Linux workstation and tried to run
6288 the Exim test suite.
6290 To get round this problem I have changed the code so that it now does its
6291 own call to the DNS to look up PTR records when searching for a host name.
6292 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6293 addresses that are only in /etc/hosts are still found.
6295 This behaviour is, however, controlled by an option called host_lookup_
6296 order, which defaults to "bydns:byaddr". If people want to use the other
6297 order, or indeed, just use one or the other means of lookup, they can
6298 specify it in this variable.
6300 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6301 some operating systems, this comes back from gethostbyaddr() as an empty
6302 string, and this is what Exim used to test for. However, it seems that in
6303 other systems, "." is yielded. Exim now tests for this case too.
6305 7. The values of check_spool_space and check_log_space are now held internally
6306 as a number of kilobytes instead of an absolute number of bytes. If a
6307 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6308 kilobyte. This means that much larger values can be stored.
6310 8. Exim monitor: an attempt to get the action menu when not actually pointing
6311 at a message produces an empty menu entitled "No message selected". This
6312 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6313 no entries in it ("Shell widget menu has zero width and/or height"). So I
6314 have added a single, blank menu entry in this case.
6316 9. Added ${quote_local_part.
6318 10. MIME decoding is now applied to the contents of Subject: header lines when
6321 11. Now that a reference to $sender_host_address automatically causes a reverse
6322 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6323 host lookup before query-style lookups in lists that might use this
6324 variable. This has therefore been abolished, and the "net-" prefix is no
6325 longer necessary for query-style lookups.
6327 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6328 have been a typo for LFLAGS, so it has been changed.
6330 13. The install script calls Exim with "-C /dev/null" in order to find the
6331 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6332 to be output. However, since Exim outputs its version number before the
6333 error, it didn't break the script. It just looked ugly. I fixed this by
6334 always allowing "-C /dev/null" if the caller is root.
6336 14. Ignore overlarge ACL variable number when reading spool file - insurance
6337 against a later release with more variables having written the file.
6339 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6340 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6341 this, as well as the form without the "IPv6" on the front.
6343 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6344 OS/Makefile-Darwin file.
6346 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6347 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6349 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6350 variable. However, some OS use ints for inodes. I've added cast to long int
6351 to get rid of the compiler warning.
6353 19. I had forgotten to lock out "/../" in configuration file names when
6354 ALT_CONFIG_PREFIX was set.
6356 20. Routers used for verification do not need to specify transports. However,
6357 if such a router generated a host list, and callout was configured, Exim
6358 crashed, because it could not find a port number from the (non-existent)
6359 transport. It now assumes port 25 in this circumstance.
6361 21. Added the -t option to exigrep.
6363 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6364 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6365 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6366 them is set, LOOKUP_LSEARCH is forced.
6368 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6370 24. Added sender and host information to the "rejected by local_scan()" log
6371 line; previously there was no indication of these.
6373 25. Added .include_if_exists.
6375 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6376 Linux. It turns out that not all file systems support this. Apparently some
6377 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6378 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6379 not supported on the file descriptor, is now ignored when Exim is trying to
6380 sync a directory. This applies only to Linux.
6382 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6384 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6385 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6388 29. Added support for maildirsize files from supplied patch (modified a bit).
6390 30. The use of :fail: followed by an empty string could lead Exim to respond to
6391 sender verification failures with (e.g.):
6393 550 Verification failed for <xxx>
6394 550 Sender verify failed
6396 where the first response line was missing the '-' that indicates it is not
6397 the final line of the response.
6399 31. The loop for finding the name of the user that called Exim had a hardwired
6400 limit of 10; it now uses the value of finduser_retries, which is used for
6401 all other user lookups.
6403 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6406 33. Exim was neglecting to zero errno before one call of strtol() when
6407 expanding a string and expecting an integer value. On some systems this
6408 resulted in spurious "integer overflow" errors. Also, it was casting the
6409 result into an int without checking.
6411 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6412 did not work. The code looks as if it has *never* worked, though it appears
6413 to have been documented since at least release 1.62. I have made it work.
6415 35. The "timeout_DNS" error in retry rules, also documented since at least
6416 1.62, also never worked. As it isn't clear exactly what this means, and
6417 clearly it isn't a major issue, I have abolished the feature by treating it
6418 as "timeout", and writing a warning to the main and panic logs.
6420 36. The display of retry rules for -brt wasn't always showing the error code
6423 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6424 timeout_connect_A, timeout_connect_MX.
6426 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6427 to the empty sender.
6429 39. The daemon was not analysing the content of -oX till after it had closed
6430 stderr and disconnected from the controlling terminal. This meant that any
6431 syntax errors were only noted on the panic log, and the return code from
6432 the command was 0. By re-arranging the code a little, I've made the
6433 decoding happen first, so such errors now appear on stderr, and the return
6434 code is 1. However, the actual setting up of the sockets still happens in
6435 the disconnected process, so errors there are still only recorded on the
6438 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6439 connections (as happens on some IP stacks) was logged at start up time as
6440 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6441 it from "IPv6 and IPv4", which means that two separate sockets are being
6444 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6445 says whether AF_INET or AF_INET6 was passed as an argument.
6447 42. Exiwhat output was messed up when time zones were included in log
6450 43. Exiwhat now gives more information about the daemon's listening ports,
6451 and whether -tls-on-connect was used.
6453 44. The "port" option of the smtp transport is now expanded.
6455 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6456 silently ignored. Now an error message is written to the main and panic
6459 46. There's a new ACL modifier called "logwrite" which writes to a log file
6460 as soon as it is encountered.
6462 47. Added $local_user_uid and $local_user_gid at routing time.
6464 48. Exim crashed when trying to verify a sender address that was being
6467 49. Exim was recognizing only a space character after ".include". It now also
6468 recognizes a tab character.
6470 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6471 extracting the relevant information from the specification. The man page no
6472 longer contains scrambled data for the -d option, and I've added a section
6473 at the front about calling Exim under different names.
6475 51. Added "extra_headers" argument to the "mail" command in filter files.
6477 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6480 53. Installed eximstats 1.29.
6482 54. Added transport_filter_timeout as a generic transport option.
6484 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6485 Cc: header lines. This was required by RFC 822, but it not required by RFC
6488 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6489 incoming messages that did not have them. Now it does so only if the
6490 message originates locally, that is, if there is no associated remote host
6491 address. When Resent- header lines are present, this applies to the Resent-
6492 lines rather than the non-Resent- lines.
6494 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6495 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6497 58. Messages for configuration errors now include the name of the main
6498 configuration file - useful now that there may be more than one file in a
6499 list (.included file names were always shown).
6501 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6502 for those rare installations that do not start the daemon as root or run it
6503 setuid root. I've cut out the call to initgroups() if the daemon is not
6506 60. The Exim user and group can now be bound into the binary as text strings
6507 that are looked up at the start of Exim's processing.
6509 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6511 62. Added $mailstore_basename variable.
6513 63. Installed patch to sieve.c from Michael Haardt.
6515 64. When Exim failed to open the panic log after failing to open the main log,
6516 the original message it was trying to log was written to stderr and debug
6517 output, but if they were not available (the usual case in production), it
6518 was lost. Now it is written to syslog before the two lines that record the
6519 failures to open the logs.
6521 65. Users' Exim filters run in subprocesses under the user's uid. It is
6522 possible for a "deliver" command or an alias in a "personal" command to
6523 provoke an address rewrite. If logging of address rewriting is configured,
6524 this fails because the process is not running as root or exim. There may be
6525 a better way of dealing with this, but for the moment (because 4.30 needs
6526 to be released), I have disabled address rewrite logging when running a
6527 filter in a non-root, non-exim process.
6533 1. The buildconfig auxiliary program wasn't quoting the value set for
6534 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6535 not defined. This bug was masked in 4.22 by the effect that was fixed in
6538 2. Some messages that were rejected after a message id was allocated were
6539 shown as "incomplete" by exigrep. It no longer does this for messages that
6540 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6542 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6543 have allow_domain_literals set, the ID did not get logged in the <= line.
6544 Domain literals are now always recognized in Message-ID: header lines.
6546 4. The first argument for a ${extract expansion item is the key name or field
6547 number. Leading and trailing spaces in this item were not being ignored,
6548 causing some misleading effects.
6550 5. When deliver_drop_privilege was set, single queue runner processes started
6551 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6552 same command in the process it spins off) were not dropping privilege.
6554 6. When the daemon running as "exim" started a queue runner, it always
6555 re-executed Exim in the spun-off process. This is a waste of effort when
6556 deliver_drop_privilege is set. The new process now just calls the
6557 queue-runner function directly.
6563 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6566 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6567 ignored. Though the use of -oP was forcing the writing of a pid file, it
6568 was always written to the default place.
6570 3. If the message "no IP address found for host xxxx" is generated during
6571 incoming verification, it is now followed by identification of the incoming
6572 connection (so you can more easily find what provoked it).
6574 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6576 5. Added some features to "harden" Exim a bit more against certain attacks:
6578 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6579 be put in Local/Makefile. This is like the never_users runtime option,
6580 but it cannot be overridden. The default setting is "root".
6582 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6583 prefix string with which any file named in a -C command line option
6586 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6587 is retained for -C and -D only if the caller of Exim is root. Without
6588 it, the exim user may also use -C and -D and retain privilege.
6590 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6591 command line option is disabled.
6593 6. Macro names set by the -D option must start with an upper case letter, just
6594 like macro names defined in the configuration file.
6596 7. Added "dereference=" facility to LDAP.
6598 8. Two instances of the typo "uknown" in the source files are fixed.
6600 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6601 the Configure-Makefile script screwed up while processing it.
6603 10. Incorporated PCRE 4.4.
6605 11. The SMTP synchronization check was not operating right at the start of an
6606 SMTP session. For example, it could not catch a HELO sent before the client
6607 waited for the greeting. There is now a check for outstanding input at the
6608 point when the greeting is written. Because of the duplex, asynchronous
6609 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6610 way, but not yet received, when the check is performed.
6612 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6613 on TCP/IP sockets, because this apparently causes some broken clients to
6616 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6617 unchanged) from the Cygwin maintainer.
6619 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6620 was supported instead of testing for mbx. Effectively a typo.
6622 15. The spa authenticator server code was not checking that the input it
6623 received was valid base64.
6625 16. The debug output line for the "set" modifier in ACLs was not showing the
6626 name of the variable that was being set.
6628 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6630 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6631 lookup on its own. Something else was needed to trigger the lookup. For
6632 example, a match in host_lookup or the need for a host name in a host list.
6633 Now, if $sender_host_name is referenced and the host name has not yet been
6634 looked up, a lookup is performed. If the lookup fails, the variable remains
6635 empty, and $host_lookup_failed is set to "1".
6637 19. Added "eqi" as a case-independent comparison operator.
6639 20. The saslauthd authentication condition could segfault if neither service
6640 nor realm was specified.
6642 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6643 error message that was logged was misleading, and incoming SMTP
6644 connections were dropped. The message is now more accurate, and temporary
6645 errors are given to SMTP connections.
6647 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6648 (except RSET) with a 421 error, until QUIT is received. However, it was
6649 failing to send a response to QUIT.
6651 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6652 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6653 ACL is now run after the helo_try_verify_hosts code.
6655 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6656 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6657 case-independent, so other case variants are also recognized). Apparently
6658 some systems use these upper case variants.
6660 25. If more than two messages were waiting for the same host, and a transport
6661 filter was specified for the transport, Exim sent two messages over the
6662 same TCP/IP connection, and then failed with "socket operation on non-
6663 socket" when it tried to send the third.
6665 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6667 27. The extern definition of crypt16() in expand.c was not being excluded when
6668 the OS had its own crypt16() function.
6670 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6671 as a preferred synonym for return_size_limit, both as an option and as an
6674 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6676 30. Changed the default configuration ACL to relax the local part checking rule
6677 for addresses that are not in any local domains. For these addresses,
6678 slashes and pipe symbols are allowed within local parts, but the sequence
6679 /../ is explicitly forbidden.
6681 31. SPA server authentication was not clearing the challenge buffer before
6684 32. log_message in a "warn" ACL statement was writing to the reject log as
6685 well as to the main log, which contradicts the documentation and doesn't
6686 seem right (because no rejection is happening). So I have stopped it.
6688 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6689 However, I am unable to do any testing of this.
6691 34. Fixed an infelicity in the appendfile transport. When checking directories
6692 for a mailbox, to see if any needed to be created, it was accidentally
6693 using path names with one or more superfluous leading slashes; tracing
6694 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6696 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6697 contents are added to the log line that is written for every discarded
6698 recipient. (Previously a log_message setting was ignored.)
6700 36. The ${quote: operator now quotes the string if it is empty.
6702 37. The install script runs exim in order to find its version number. If for
6703 some reason other than non-existence or emptiness, which it checks, it
6704 could not run './exim', it was installing it with an empty version number,
6705 i.e. as "exim-". This error state is now caught, and the installation is
6708 38. An argument was missing from the function that creates an error message
6709 when Exim fails to connect to the socket for saslauthd authentication.
6710 This could cause Exim to crash, or give a corrupted message.
6712 39. Added isip, isip4, and isip6 to ${if conditions.
6714 40. The ACL variables $acl_xx are now saved with the message, and can be
6715 accessed later in routers, transports, and filters.
6717 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6718 strings in the file are not string-expanded.
6720 42. If a MAIL command specified a SIZE value that was too large to fit into an
6721 int variable, the check against message_size_limit failed. Such values are
6722 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6723 day this will have to be increased, but I don't think I want to be around
6724 when emails are that large.
6731 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6732 iconv() is not standard in FreeBSD.
6734 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6735 IPv6 enabled. The observed symptom was a segmentation fault on return from
6736 the function os_common_find_running_interfaces() in src/os.c.
6738 3. In the check_special_case() function in daemon.c I had used "errno" as an
6739 argument name, which causes warnings on some systems. This was basically a
6740 typo, since it was named "eno" in the comments!
6742 4. The code that waits for the clock to tick (at a resolution of some fraction
6743 of a second) so as to ensure message-id uniqueness was always waiting for
6744 at least one whole tick, when it could have waited for less. [This is
6745 almost certainly not relevant at current processor speeds, where it is
6746 unlikely to ever wait at all. But we try to future-proof.]
6748 5. The function that sleeps for a time interval that includes fractions of a
6749 second contained a race. It did not block SIGALRM between setting the
6750 timer, and suspending (a couple of lines later). If the interval was short
6751 and the sigsuspend() was delayed until after it had expired, the suspension
6752 never ended. On busy systems this could lead to processes getting stuck for
6755 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6756 process, before it forks any delivery processes. The open lookup caching
6757 mechanism meant that the open file or database connection was passed into
6758 the delivery process. The problem was that delivery processes always tidy
6759 up cached lookup data. This could cause a problem for the next delivery
6760 process started by the queue runner, because the external queue runner
6761 process does not know about the closure. So the next delivery process
6762 still has data in the lookup cache. In the case of a file lookup, there was
6763 no problem because closing a file descriptor in a subprocess doesn't affect
6764 the parent. However, if the lookup was caching a connection to a database,
6765 the connection was closed, and the second delivery process was likely to
6766 see errors such as "PGSQL: query failed: server closed the connection
6767 unexpectedly". The problem has been fixed by closing all cached lookups
6768 in a queue runner before running a delivery process.
6770 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6771 seem to have the "const" qualifier which it has on other OS. I've
6774 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6775 *of the same type* (client or server) with the same public name that an
6776 error should be diagnosed.
6778 9. When Exim looked up a host name for an IP address, but failed to find the
6779 original IP address when looking up the host name (a safety check), it
6780 output the message "<ip address> does not match any IP for NULL", which was
6781 confusing, to say the least. The bug was that the host name should have
6782 appeared instead of "NULL".
6784 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6785 user that is built into the binary, and the -C or -D options is used, root
6786 privilege is dropped before the configuration file is read. In addition,
6787 logging is switched to stderr instead of the normal log files. If the
6788 configuration then re-defines the Exim user, the unprivileged environment
6789 is probably not what is expected, so Exim logs a panic warning message (but
6792 However, if deliver_drop_privilege is set, the unprivileged state may well
6793 be exactly what is intended, so the warning has been cut out in that case,
6794 and Exim is allowed to try to write to its normal log files.
6800 1. smtp_return_error_details was not giving details for temporary sender
6801 or receiver verification errors.
6803 2. Diagnose a configuration error if two authenticators have the same public
6806 3. Exim used not to create the message log file for a message until the first
6807 delivery attempt. This could be confusing when incoming messages were held
6808 for policy or load reasons. The message log file is now created at the time
6809 the message is received, and an initial "Received" line is written to it.
6811 4. The automatically generated man page for command line options had a minor
6812 bug that caused no ill effects; however, a more serious problem was that
6813 the procedure for building the man page automatically didn't always
6814 operate. Consequently, release 4.20 contains an out-of-date version. This
6815 shouldn't happen again.
6817 5. When building Exim with embedded Perl support, the script that builds the
6818 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6819 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6821 6. The freeze_tell option was not being used for messages that were frozen on
6822 arrival, either by an ACL or by local_scan().
6824 7. Added the smtp_incomplete_transaction log selector.
6826 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6827 was accepting AUTH without a new EHLO.
6829 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6830 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6833 10. Logging of TCP/IP connections (when configured) now happens in the main
6834 daemon process instead of the child process, so that the TCP/IP connection
6835 count is more accurate (but it can never be perfect).
6837 11. The use of "drop" in a nested ACL was not being handled correctly in the
6838 outer ACL. Now, if condition failure induced by the nested "drop" causes
6839 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6840 or "require"), the connection is dropped.
6842 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6843 that yield "discard" can now be used with an "accept" or a "discard" verb,
6844 but an error is generated for any others (because I can't see a useful way
6845 to define what should happen).
6847 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6848 are now processed in the same way as lines in the Exim configuration file.
6849 In particular, continuation lines are supported.
6851 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6853 15. Added -ti meaning -t -i.
6855 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6856 domains, and warn by logging if others are found.
6858 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6859 returned from GIFCONF when Exim is trying to find the list of interfaces on
6860 a host. The code in os.c has been modified to copy each ifreq to an aligned
6861 structure in all cases.
6863 Also, in some cases, the returned ifreq's were being copied to a 'struct
6864 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6865 means the last couple of bytes of an IPv6 address could be chopped if the
6866 ifreq contained only a normal sockaddr (14 bytes storage).
6868 18. Named domain lists were not supported in the hosts_treat_as_local option.
6869 An entry such as +xxxx was not recognized, and was treated as a literal
6872 19. Ensure that header lines added by a DATA ACL are included in the reject log
6873 if the ACL subsequently rejects the message.
6875 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6876 MD5 (which is deprecated).
6878 21. When testing a filter file using -bf, Exim was writing a message when it
6879 took the sender from a "From " line in the message, but it was not doing so
6880 when it took $return_path from a Return-Path: header line. It now does.
6882 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6883 with a valid header line field name (a series of printing characters
6884 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6886 23. Changed "disc" in the source to "disk" to conform to the documentation and
6887 the book and for uniformity.
6889 24. Ignore Sendmail's -Ooption=value command line item.
6891 25. When execve() failed while trying to run a command in a pipe transport,
6892 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6893 could be confused with a return value of 69 from the command itself. This
6894 has been changed to 127, the value the shell returns if it is asked to run
6895 a non-existent command. The wording for the related log line suggests a
6896 non-existent command as the problem.
6898 26. If received_header_text expands to an empty string, do not add a Received:
6899 header line to the message. (Well, it adds a token one on the spool, but
6900 marks it "old" so that it doesn't get used or transmitted.)
6902 27. Installed eximstats 1.28 (addition of -nt option).
6904 28. There was no check for failure on the call to getsockname() in the daemon
6905 code. This can fail if there is a shortage of resources on the system, with
6906 ENOMEM, for example. A temporary error is now given on failure.
6908 29. Contrary to the C standard, it seems that in some environments, the
6909 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6910 program. Exim now does this explicitly; it affects the formatting of
6911 timestamps using strftime().
6913 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6914 complaints. I've now initialized all the variables, to avoid this.
6916 32. Header lines added by a system filter were not being "seen" during
6917 transport-time rewrites.
6919 33. The info_callback() function passed to OpenSSL is set up with type void
6920 (*)(SSL *, int, int), as described somewhere. However, when calling the
6921 function (actually a macro) that sets it up, the type void(*)() is
6922 expected. I've put in a cast to prevent warnings from picky compilers.
6924 34. If a DNS black list lookup found a CNAME record, but there were no A
6925 records associated with the domain it pointed at, Exim crashed.
6927 35. If a DNS black list lookup returned more than one A record, Exim ignored
6928 all but the first. It now scans all returned addresses if a particular IP
6929 value is being sought. In this situation, the contents of the
6930 $dnslist_value variable are a list of all the addresses, separated by a
6933 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6934 to accept a host name and all its aliases if the forward lookup for any of
6935 them yielded the IP address of the incoming connection. Now it accepts only
6936 those names whose forward lookup yields the correct IP address. Any other
6937 names are discarded. This closes a loophole whereby a rogue DNS
6938 administrator could create reverse DNS records to break through a
6939 wildcarded host restriction in an ACL.
6941 37. If a user filter or a system filter that ran in a subprocess used any of
6942 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6943 the wrong values were passed to the pipe command ($thisaddress had the
6944 value of $0, $0 had the value of $1, etc). This bug was introduced by
6945 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6947 38. Improved the line breaking for long SMTP error messages from ACLs.
6948 Previously, if there was no break point between 40 and 75 characters, Exim
6949 left the rest of the message alone. Two changes have been made: (a) I've
6950 reduced the minimum length to 35 characters; (b) if it can't find a break
6951 point between 35 and 75 characters, it looks ahead and uses the first one
6952 that it finds. This may give the occasional overlong line, but at least the
6953 remaining text gets split now.
6955 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6956 file descriptors might be low, and that setting 1000 would always raise it.
6957 It turns out that in some environments, the limit is already over 1000 and
6958 that lowering it causes trouble. So now Exim takes care not to decrease it.
6960 40. When delivering a message, the value of $return_path is set to $sender_
6961 address at the start of routing (routers may change the value). By an
6962 oversight, this default was not being set up when an address was tested by
6963 -bt or -bv, which affected the outcome if any router or filter referred to
6966 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6967 log only when "message" or "log_message" are set. However, if one of the
6968 conditions was an address verification, or a call to a nested ACL, the
6969 messages generated by the underlying test were being passed through. This
6970 no longer happens. The underlying message is available in $acl_verify_
6971 message for both "message" and "log_message" expansions, so it can be
6972 passed through if needed.
6974 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6975 new expansion $bh_ to give the encoded byte string without charset
6976 translation. Translation happens only if iconv() is available; HAVE_ICONV
6977 indicates this at build time. HEADERS_CHARSET gives the charset to
6978 translate to; headers_charset can change it in the configuration, and
6979 "headers charset" can change it in an individual filter file.
6981 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6982 that creates RFC 2047 encoded "words" labels them as that charset instead
6983 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6984 expansion operator; (ii) when Exim creates a From: line for a local
6985 message; (iii) when a header line is rewritten to include a "phrase" part.
6987 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6988 buggy, causing it to skip the first lines of messages whose message ID
6989 ended in 'D'. This would not have bitten before Exim release 4.14, because
6990 message IDs were unlikely to end in 'D' before then. The effect was to have
6991 incorrect size information for certain domains.
6993 45. #include "config.h" was missing at the start of the crypt16.c module. This
6994 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6997 46. If there was a timeout during a "random" callout check, Exim treated it as
6998 a failure of the random address, and carried on sending RSET and the real
6999 address. If the delay was just some slowness somewhere, the response to the
7000 original RCPT would be taken as a response to RSET and so on, causing
7001 mayhem of various kinds.
7003 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7004 when I implemented it. It didn't allow for the fact that some option values
7005 may legitimately be negative (e.g. size_addition), and it didn't even do
7006 the right test for positive values.
7008 48. Domain names in DNS records are case-independent. Exim always looks them up
7009 in lower case. Some resolvers return domain names in exactly the case they
7010 appear in the zone file, that is, they may contain uppercase letters. Not
7011 all resolvers do this - some return always lower case. Exim was treating a
7012 change of case by a resolver as a change of domain, similar to a widening
7013 of a domain abbreviation. This triggered its re-routing code and so it was
7014 trying to route what was effectively the same domain again. This normally
7015 caused routing to fail (because the router wouldn't handle the domain
7016 twice). Now Exim checks for this case specially, and just changes the
7017 casing of the domain that it ultimately uses when it transmits the message
7020 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7023 50. If a filter generated a file delivery with a non-absolute name (possible if
7024 no home directory exists for the router), the forbid_file option was not
7027 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7028 the existing equality matching.
7030 52. Exim was using ints instead of ino_t variables in some places where it was
7031 dealing with inode numbers.
7033 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7034 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7035 called TMPDIR, and if it finds it is different, it changes its value.
7037 54. The smtp_printf() function is now made available to local_scan() so
7038 additional output lines can be written before returning. There is also an
7039 smtp_fflush() function to enable the detection of a dropped connection.
7040 The variables smtp_input and smtp_batched_input are exported to
7043 55. Changed the default runtime configuration: the message "Unknown user"
7044 has been removed from the ACL, and instead placed on the localuser router,
7045 using the cannot_route_message feature. This means that any verification
7046 failures that generate their own messages won't get overridden. Similarly,
7047 the "Unrouteable address" message that was in the ACL for unverifiable
7048 relay addresses has also been removed.
7050 56. Added hosts_avoid_esmtp to the smtp transport.
7052 57. The exicyclog script was not checking for the esoteric option
7053 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7054 will work only if exicyclog is run under the appropriate euid.
7056 58. Following a discussion on the list, the rules by which Exim recognises line
7057 endings on incoming messages have been changed. The -dropcr and drop_cr
7058 options are now no-ops, retained only for backwards compatibility. The
7059 following line terminators are recognized: LF CRLF CR. However, special
7060 processing applies to CR:
7062 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7063 nor a local message in the state where . is a terminator.
7065 (ii) If a bare CR is encountered in a header line, an extra space is added
7066 after the line terminator so as not to end the header. The reasoning
7067 behind this is that bare CRs in header lines are most likely either
7068 to be mistakes, or people trying to play silly games.
7070 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7071 was being incorrectly given as 18 bytes larger than it should have been.
7072 This is a VOB (very old bug).
7074 60. This may never have affected anything current, but just in case it has:
7075 When the local host is found other than at the start of a list of hosts,
7076 the local host, those with the same MX, and any that follow, are discarded.
7077 When the list in question was part of a longer list of hosts, the following
7078 hosts (not currently being processed) were also being discarded. This no
7079 longer happens. I'm not sure if this situation could ever has previously
7082 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7085 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7086 header. This is something that is recommended in a new Internet Draft, and
7087 is something that is documented as being done by Sendmail. There are two
7088 possible values. For messages generated by the autoreply transport, Exim
7091 Auto-Submitted: auto-replied
7093 whereas for all other generated messages (e.g. bounces) it adds
7095 Auto-Submitted: auto-generated
7097 63. The "personal" condition in filters now includes a test for the
7098 Auto-Submitted: header. If it contains the string "auto-" the message it
7099 not considered personal.
7101 64. Added rcpt_include_affixes as a generic transport option.
7103 65. Added queue_only_override (default true).
7105 66. Added the syslog_duplication option.
7107 67. If what should have been the first header line of a message consisted of
7108 a space followed by a colon, Exim was mis-interpreting it as a header line.
7109 It isn't of course - it is syntactically invalid and should therefore be
7110 treated as the start of the message body. The misbehaviour could have
7111 caused a number of strange effects, including loss of data in subsequent
7112 header lines, and spool format errors.
7114 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7115 client host had authenticated. This control can now be exercised by an ACL
7116 for more flexibility.
7118 69. By default, callouts do not happen when testing with -bh. There is now a
7119 variant, -bhc, which does actually run the callout code, including
7120 consulting and updating the callout cache.
7122 70. Added support for saslauthd authentication, courtesy of Alexander
7125 71. If statvfs() failed on the spool or log directories while checking their
7126 size for availability, Exim confusingly gave the error "space shortage".
7127 Furthermore, in debugging mode it crashed with a floating point exception.
7128 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7129 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7130 serious problem, Exim now writes to the main and panic logs when this
7131 happens, with details of the failure. It then refuses to accept the
7132 incoming message, giving the message "spool directory problem" or "log
7133 directory problem" with a 421 code for SMTP messages.
7135 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7136 0, 1, and 2 exist, because some OS complain for execs without them (see
7137 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7138 descriptors. However, the code omitted to check that the open succeeded,
7139 causing mysterious errors if for some reason the permissions on /dev/null
7140 got screwed. Now Exim writes a message to the main and panic logs, and
7141 bombs out if it can't open /dev/null.
7143 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7144 interact so that it is all more flexible. It is supposed to remain
7145 backwards compatible. Also added extra_local_interfaces.
7147 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7148 to bomb out with an assertion failure - to the client this appears as a
7149 connection drop. This problem occurs in the part of the code that was taken
7150 from the Samba project. Fortunately, the assertion is in a very simple
7151 function, so I have fixed this by reproducing the function inline in the
7152 one place where it is called, and arranging for authentication to fail
7153 instead of killing the process with assert().
7155 75. The SPA client code was not working when the server requested OEM rather
7156 than Unicode encoding.
7158 76. Added code to make require_files with a specific uid setting more usable in
7159 the case where statting the file as root fails - usually a non-root-mounted
7160 NFS file system. When this happens and the failure is EACCES, Exim now
7161 forks a subprocess and does the per-uid checking as the relevant uid.
7163 77. Added process_log_path.
7165 78. If log_file_path was not explicitly set, a setting of check_log_space or
7166 check_log_inodes was ignored.
7168 79. If a space check for the spool or log partitions fails, the incident is now
7169 logged. Of course, in the latter case the data may get lost...
7171 80. Added the %p formatting code to string_format() so that it can be used to
7172 print addresses in debug_print(). Adjusted all the address printing in the
7173 debugging in store.c to use %p rather than %d.
7175 81. There was a concern that a line of code in smtp_in.c could overflow a
7176 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7177 initially expressed, the concern was not well-founded, because trailing
7178 spaces are removed early. However, if the trailing spaces were followed by
7179 a NULL, they did not get removed, so the overflow was possible. Two fixes
7182 (a) I re-wrote the offending code in a cleaner fashion.
7183 (b) If an incoming SMTP command contains a NULL character, it is rejected
7186 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7187 runs initgroups(), so that if the Exim user is in any additional groups,
7188 they will be used during message reception.
7194 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob