1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
18 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
19 When redis_servers is set to a list of > 1 element, and the Redis servers
20 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
21 case of MOVED into a DEFER case instead, thus moving the query onto the
22 next server in the list. For a cluster of N elements, all N servers must
23 be defined in redis_servers.
25 GF/02 Catch and remove uninitialized value warning in exiqsumm
26 Check for existence of @ARGV before looking at $ARGV[0]
28 JH/01 Replace the store_release() internal interface with store_newblock(),
29 which internalises the check required to safely use the old one, plus
30 the allocate and data copy operations duplicated in both (!) of the
33 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
34 modifier. This matches the restriction on the commandline.
36 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
37 Previously only the last row was returned.
39 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
40 we assumed that tags in the header were well-formed, and parsed the
41 element content after inspecting only the first char of the tag.
42 Assumptions at that stage could crash the receive process on malformed
45 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
46 While running the DKIM ACL we operate on the Permanent memory pool so that
47 variables created with "set" persist to the DATA ACL. Also (at any time)
48 DNS lookups that fail create cache records using the Permanent pool. But
49 expansions release any allocations made on the current pool - so a dnsdb
50 lookup expansion done in the DKIM ACL releases the memory used for the
51 DNS negative-cache, and bad things result. Solution is to switch to the
52 Main pool for expansions.
53 While we're in that code, add checks on the DNS cache during store_reset,
54 active in the testsuite.
55 Problem spotted, and debugging aided, by Wolfgang Breyha.
57 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
58 When none of the hosts presented to a transport match an already-open
59 connection, close it and proceed with the list. Previously we would
60 queue the message. Spotted by Lena with Yahoo, probably involving
63 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
64 Previously a spurious "250 OK id=" response was appended to the proper
67 JH/08 The "support for" informational output now, which built with Content
68 Scanning support, has a line for the malware scanner interfaces compiled
69 in. Interface can be individually included or not at build time.
71 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
72 by the template makefile "src/EDITME". The "STREAM" support for an older
73 ClamAV interface method is removed.
75 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
76 rows affected is given instead).
78 JH/11 The runtime Berkeley DB library version is now additionally output by
79 "exim -d -bV". Previously only the compile-time version was shown.
81 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
82 SMTP connection. Previously, when one had more receipients than the
83 first, an abortive onward connection was made. Move to full support for
84 multiple onward connections in sequence, handling cutthrough connection
85 for all multi-message initiating connections.
87 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
88 routers. Previously, a multi-recipient message would fail to match the
89 onward-connection opened for the first recipient, and cause its closure.
91 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
92 a timeout on read on a GnuTLS initiating connection, resulting in the
93 initiating connection being dropped. This mattered most when the callout
94 was marked defer_ok. Fix to keep the two timeout-detection methods
97 JH/15 Relax results from ACL control request to enable cutthrough, in
98 unsupported situations, from error to silently (except under debug)
99 ignoring. This covers use with PRDR, frozen messages, queue-only and
102 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
104 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
105 metadata, resulting in a crash in free().
107 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
108 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
109 Broken also in d185889f4, with init system revamp.
111 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
112 Previously we abruptly closed the connection after reading a malware-
113 found indication; now we go on to read the "scan ok" response line,
116 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
117 ACL. Previously, a crash would result.
119 JH/19 Speed up macro lookups during configuration file read, by skipping non-
120 macro text after a replacement (previously it was only once per line) and
121 by skipping builtin macros when searching for an uppercase lead character.
123 JH/20 DANE support moved from Experimental to mainline. The Makefile control
124 for the build is renamed.
126 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
127 was allocated for every new TLS startup, meaning one per message. Fix
128 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
130 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
131 reported the original. Fix to report (as far as possible) the ACL
132 result replacing the original.
134 JH/23 Fix memory leak during multi-message connections using STARTTLS under
135 OpenSSL. Certificate information is loaded for every new TLS startup,
136 and the resources needed to be freed.
138 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
140 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
143 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
144 DATA response info to the (existing) per-recipient response info for
145 the "C=" log element. It can have useful tracking info from the
146 destination system. Patch from Simon Arlott.
148 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
149 length value. Previously this would segfault.
151 HS/02 Support Avast multiline protoocol, this allows passing flags to
152 newer versions of the scanner.
154 JH/28 Ensure that variables possibly set during message acceptance are marked
155 dead before release of memory in the daemon loop. This stops complaints
156 about them when the debug_store option is enabled. Discovered specifically
157 for sender_rate_period, but applies to a whole set of variables.
158 Do the same for the queue-runner and queue-list loops, for variables set
159 from spool message files. Do the same for the SMTP per-message loop, for
160 certain variables indirectly set in ACL operations.
162 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
163 as a multi-recipient message from a mailinglist manager). The coding had
164 an arbitrary cutoff number of characters while checking for more input;
165 enforced by writing a NUL into the buffer. This corrupted long / fast
166 input. The problem was exposed more widely when more pipelineing of SMTP
167 responses was introduced, and one Exim system was feeding another.
168 The symptom is log complaints of SMTP syntax error (NUL chars) on the
169 receiving system, and refused recipients seen by the sending system
170 (propating to people being dropped from mailing lists).
171 Discovered and pinpointed by David Carter.
173 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
174 replaced by the ${authresults } expansion.
176 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
178 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
179 allows proper process termination in container environments.
181 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
182 Previously the "final dot" had a newline after it; ensure it is CR,LF.
184 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
185 and "err_perm", deprecated since 4.83 when the RFC-defined words
186 "temperror" and "permerror" were introduced.
188 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
189 transport-filters or DKIM-signing. The restriction was lost in the
190 consolidation of verify-callout and delivery SMTP handling.
191 Extend the restriction to also cover ARC-signing.
193 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
194 in defer=pass mode supply a 450 to the initiator. Previously the message
197 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
198 tls_require_ciphers is used as before.
200 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
201 "pass_unscanned". Only tmpfails from the scanner are written to
202 the paniclog, as they may require admin intervention (permission
203 denied, license issues). Other scanner errors (like decompression
204 bombs) do not cause a paniclog entry.
206 JH/36 Fix reinitialisation of DKIM logging variable between messages.
207 Previously it was possible to log spurious information in receive log
210 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
211 triggered odd behaviour from Outlook Express clients.
213 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
216 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
217 since the IETF WG has not yet settled on that versus the original
218 "bare" representation.
220 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
221 Previously the millisecond value corrupted the output.
222 Fix also for syslog_pid=no and log_selector +pid, for which the pid
223 corrupted the output.
229 JH/01 Rework error string handling in TLS interface so that the caller in
230 more cases is responsible for logging. This permits library-sourced
231 string to be attached to addresses during delivery, and collapses
232 pairs of long lines into single ones.
234 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
235 during configuration. Wildcards are allowed and expanded.
237 JH/02 Rework error string handling in DKIM to pass more info back to callers.
238 This permits better logging.
240 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
241 do not close it down and have the child transport start it up again on
242 the passed-on TCP connection. Instead, proxy the child (and any
243 subsequent ones) for TLS via a unix-domain socket channel. Logging is
244 affected: the continued delivery log lines do not have any DNSSEC, TLS
245 Certificate or OCSP information. TLS cipher information is still logged.
247 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
248 identical IP addresses on different listening ports. Will also affect
251 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
252 add noisy ifdef guards to special-case this sillyness.
253 Patch from Bernd Kuhls.
255 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
256 than 255 are no longer allowed.
258 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
259 Disable the session-cache too, which might reduce our load. Since we
260 currrectly use a new context for every connection, both as server and
261 client, there is no benefit for these.
262 GnuTLS appears to not support tickets server-side by default (we don't
263 call gnutls_session_ticket_enable_server()) but client side is enabled
264 by default on recent versions (3.1.3 +) unless the PFS priority string
267 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
268 <https://reproducible-builds.org/specs/source-date-epoch/>.
270 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
271 the check for any unsuccessful recipients did not notice the limit, and
272 erroneously found still-pending ones.
274 JH/08 Pipeline CHUNKING command and data together, on kernels that support
275 MSG_MORE. Only in-clear (not on TLS connections).
277 JH/09 Avoid using a temporary file during transport using dkim. Unless a
278 transport-filter is involved we can buffer the headers in memory for
279 creating the signature, and read the spool data file once for the
280 signature and again for transmission.
282 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
283 4.77 as the kernel support then wasn't solid, having issues in 64bit
284 mode. Now, it's been long enough. Add support for FreeBSD also.
286 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
287 case where the routing stage had gathered several addresses to send to
288 a host before calling the transport for the first, we previously failed
289 to close down TLS in the old transport process before passing the TCP
290 connection to the new process. The new one sent a STARTTLS command
291 which naturally failed, giving a failed delivery and bloating the retry
292 database. Investigation and fix prototype from Wolfgang Breyha.
294 JH/12 Fix check on SMTP command input synchronisation. Previously there were
295 false-negatives in the check that the sender had not preempted a response
296 or prompt from Exim (running as a server), due to that code's lack of
297 awareness of the SMTP input buffering.
299 PP/04 Add commandline_checks_require_admin option.
300 Exim drops privileges sanely, various checks such as -be aren't a
301 security problem, as long as you trust local users with access to their
302 own account. When invoked by services which pass untrusted data to
303 Exim, this might be an issue. Set this option in main configuration
304 AND make fixes to the calling application, such as using `--` to stop
307 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
308 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
309 responses to those, into a single TLS record each way (this usually means
310 a single packet). As a side issue, smtp_enforce_sync now works on TLS
313 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
314 affects you only if you're dancing at the edge of the param size limits.
315 If you are, and this message makes sense to you, then: raise the
316 configured limit or use OpenSSL 1.1. Nothing we can do for older
319 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
320 cmdline element to get the documented default one. Previously it was
323 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
324 are unsuable for non-message-specific reasons - eg. connection timeout,
325 banner-time rejection.
327 JH/16 Fix logging of delivery remote port, when specified by router, under
330 PP/06 Repair manualroute's ability to take options in any order, even if one
331 is the name of a transport.
334 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
336 JH/17 Change the list-building routines interface to use the expanding-string
337 triplet model, for better allocation and copying behaviour.
339 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
340 Previously it was constructed the first time a possibly-matching string
341 was met in the configuration file input during startup; now it is done
344 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
345 compatible one, to avoid the (poorly documented) possibility of a config
346 file in the working directory redirecting the DB files, possibly correpting
347 some existing file. CVE-2017-10140 assigned for BDB.
349 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
350 cache-hot. Previously, although the result was properly cached, the
351 initial verify call returned a defer.
353 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
354 the main verify for receipient in uncached-mode.
356 JH/22 Retire historical build files to an "unsupported" subdir. These are
357 defined as "ones for which we have no current evidence of testing".
359 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
360 if present. Previously it was ignored.
362 JH/24 Start using specified-initialisers in C structure init coding. This is
363 a C99 feature (it's 2017, so now considered safe).
365 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
366 if was a fixed-sized field and bitmask ops via macros; it is now more
369 PP/07 GitHub PR 56: Apply MariaDB build fix.
370 Patch provided by Jaroslav Škarvada.
372 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
373 during Coverity cleanups [4.87 JH/47]
374 Diagnosis and fix provided by Michael Fischer v. Mollard.
376 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
377 the right size to place the terminating semicolon on its own folded
378 line, the header hash was calculated to an incorrect value thanks to
379 the (relaxed) space the fold became.
381 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
382 and confused the parent.
384 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
385 which could crash as a result. This could lead to undeliverable messages.
387 JH/28 Logging: "next input sent too soon" now shows where input was truncated
390 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
391 matters on fast-turnover and PID-randomising systems, which were getting
392 out-of-order delivery.
394 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
395 a possibly-overlapping copy. The symptom was that "Remote host closed
396 connection in response to HELO" was logged instead of the actual 4xx
399 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
400 Previously only that bufferd was discarded, resulting in SYMTP command
403 JH/32 DKIM: when a message has multiple signatures matching an identity given
404 in dkim_verify_signers, run the dkim acl once for each. Previously only
405 one run was done. Bug 2189.
407 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
408 "panic the current process" to "deliberately defer". The panic log is
409 still written with the problem list name; the mail and reject logs now
410 get a temp-reject line for the message that was being handled, saying
411 something like "domains check lookup or other defer". The SMTP 451
412 message is still "Temporary local problem".
414 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
415 A crafted sequence of BDAT commands could result in in-use memory beeing
416 freed. CVE-2017-16943.
418 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
419 from SMTP input. Previously it was always done; now only done for DATA
420 and not BDAT commands. CVE-2017-16944.
422 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
423 to the message (such as an overlong header line). Previously this was
424 not done and we did not exit BDAT mode. Followon from the previous item
425 though a different problem.
431 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
432 than -2003 did; needs libidn2 in addition to libidn.
434 JH/02 The path option on a pipe transport is now expanded before use.
436 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
437 Patch provided by "Björn", documentation fix added too.
439 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
440 missing a wire-to-host endian conversion.
442 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
443 close after a BDAT command line could be taken as a following command,
444 giving a synch failure. Fix by only checking for synch immediately
445 before acknowledging the chunk.
447 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
448 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
449 macro. Patches provided by Josh Soref.
451 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
452 Previously we did not; the RFC seems ambiguous and VRFY is not listed
453 by IANA as a service extension. However, John Klensin suggests that we
456 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
457 the dkim code may be unix-mode line endings rather than smtp wire-format
458 CRLF, so prepend a CR to any bare LF.
460 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
461 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
463 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
464 the first were themselves being wrongly included in the feed into dkim
465 processing; with most chunk sizes in use this resulted in an incorrect
466 body hash calculated value.
468 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
469 DKIM signature block, for verification. Although advised against by
470 standards it is specifically not ruled illegal.
472 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
474 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
475 missing a body hash (the bh= tag).
477 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
478 It seems that HAProxy sends the Proxy Protocol information in clear and
479 only then does a TLS startup, so do the same.
481 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
482 TCP connections (such as for Spamd) unless the daemon successfully set
483 Fast Open mode on its listening sockets. This fixes breakage seen on
484 too-old kernels or those not configured for Fast Open, at the cost of
485 requiring both directions being enabled for TFO, and TFO never being used
486 by non-daemon-related Exim processes.
488 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
489 endings, at least on the first header line. Try to canonify any that get
490 past that check, despite the cost.
492 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
493 now limited to an arbitrary five deep, while parsing addresses with the
494 strip_excess_angle_brackets option enabled.
496 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
497 instead leave the unprompted TLS handshake in socket buffer for the
498 TLS library to consume.
500 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
502 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
504 JH/16 Drop variables when they go out of scope. Memory management drops a whole
505 region in one operation, for speed, and this leaves assigned pointers
506 dangling. Add checks run only under the testsuite which checks all
507 variables at a store-reset and panics on a dangling pointer; add code
508 explicitly nulling out all the variables discovered. Fixes one known
509 bug: a transport crash, where a dangling pointer for $sending_ip_address
510 originally assigned in a verify callout, is re-used.
512 PP/06 Drop '.' from @INC in various Perl scripts.
514 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
516 PP/08 Reduce a number of compilation warnings under clang; building with
517 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
518 should be warning-free.
520 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
522 HS/01 Fix portability problems introduced by PP/08 for platforms where
523 realloc(NULL) is not equivalent to malloc() [SunOS et al].
525 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
526 chunk. This allows us to accept broken chunked messages. We need a more
527 general solution here.
529 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
530 already-broken messages in the queue.
532 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
534 JH/19 Fix reference counting bug in routing-generated-address tracking.
540 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
541 supports it and a size is available (ie. the sending peer gave us one).
543 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
544 years of being deprecated). The replacements are the ACLs
545 acl_smtp_mime and acl_not_smtp_mime.
547 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
548 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
549 or even an in-clear connection were permitted. Now, if the host lookup
550 was dnssec and dane was requested then the host is only used if the
551 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
552 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
553 if one fails this test.
554 This means that a poorly-configured remote DNS will make it incommunicado;
555 but it protects against a DNS-interception attack on it.
557 JH/04 Bug 1810: make continued-use of an open smtp transport connection
558 non-noisy when a race steals the message being considered.
560 JH/05 If main configuration option tls_certificate is unset, generate a
561 self-signed certificate for inbound TLS connections.
563 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
564 in rewrites and routers.
566 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
567 and logged a warning sing 4.83; now they are a configuration file error.
569 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
570 (lacking @domain). Apply the same qualification processing as RCPT.
572 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
574 JH/10 Support ${sha256:} applied to a string (as well as the previous
577 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
578 a cutthrough deliver is pending, as we always want to make a connection.
579 This also avoids re-routing the message when later placing the cutthrough
580 connection after a verify cache hit.
581 Do not update it with the verify result either.
583 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
584 when routing results in more than one destination address.
586 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
587 signing (which inhibits the cutthrough capability). Previously only
588 the presence of an option was tested; now an expansion evaluating as
589 empty is permissible (obviously it should depend only on data available
590 when the cutthrough connection is made).
592 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
593 the relevant preceding SMTP command did not note the pipelining mode.
595 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
596 Previously they were not counted.
598 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
599 as one having no matching records. Previously we deferred the message
600 that needed the lookup.
602 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
603 distinguished as "(=".
605 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
606 for missing MX records. Previously it only worked for missing A records.
608 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
610 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
611 after the data-go-ahead and data-ack. Patch from Jason Betts.
613 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
614 even for a "none" policy. Patch from Tony Meyer.
616 JH/22 Fix continued use of a connection for further deliveries. If a port was
617 specified by a router, it must also match for the delivery to be
620 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
621 When one of the recipients of a message was unsuitable for the connection
622 (has no matching addresses), we lost track of needing to mark it
623 deferred. As a result mail would be lost.
625 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
627 JH/25 Decoding ACL controls is now done using a binary search; the source code
628 takes up less space and should be simpler to maintain. Merge the ACL
629 condition decode tables also, with similar effect.
631 JH/26 Fix problem with one_time used on a redirect router which returned the
632 parent address unchanged. A retry would see the parent address marked as
633 delivered, so not attempt the (identical) child. As a result mail would
636 JH/27 Fix a possible security hole, wherein a process operating with the Exim
637 UID can gain a root shell. Credit to http://www.halfdog.net/ for
638 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
641 JH/28 Enable {spool,log} filesystem space and inode checks as default.
642 Main config options check_{log,spool}_{inodes,space} are now
643 100 inodes, 10MB unless set otherwise in the configuration.
645 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
646 Previously it only applied to the main-section connection policy
649 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
651 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
652 by me. Added RFC7919 DH primes as an alternative.
654 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
655 are not in the system include path.
657 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
658 GnuTLS, when a session startup failed (eg because the client disconnected)
659 Exim did stdio operations after fclose. This was exposed by a recent
660 change which nulled out the file handle after the fclose.
662 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
663 signed directly by the cert-signing cert, rather than an intermediate
664 OCSP-signing cert. This is the model used by LetsEncrypt.
666 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
668 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
669 an incoming connection.
671 HS/02 Bug 1802: Do not half-close the connection after sending a request
674 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
675 fallback to "prime256v1".
677 JH/34 SECURITY: Use proper copy of DATA command in error message.
678 Could leak key material. Remotely exploitable. CVE-2016-9963.
684 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
685 and 3.4.4 - once the server is enabled to respond to an OCSP request
686 it does even when not requested, resulting in a stapling non-aware
687 client dropping the TLS connection.
689 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
690 support variable-length bit vectors. No functional change.
692 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
693 The I= interface field on outgoing lines is now after the H= remote
694 host field, same as incoming lines. There is a separate
695 outgoing_interface log selector which allows you to disable the
698 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
699 If not running log_selector +smtp_connection the mainlog would be held
700 open indefinitely after a "too many connections" event, including to a
701 deleted file after a log rotate. Leave the per net connection logging
702 leaving it open for efficiency as that will be quickly detected by the
703 check on the next write.
705 HS/01 Bug 1671: Fix post transport crash.
706 Processing the wait-<transport> messages could crash the delivery
707 process if the message IDs didn't exist for some reason. When
708 using 'split_spool_directory=yes' the construction of the spool
709 file name failed already, exposing the same netto behaviour.
711 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
712 mime_regex ACL conditions.
714 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
715 to DSN fail messages (bounces): remote IP, remote greeting, remote response
716 to HELO, local diagnostic string.
718 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
719 log line to debug. Even when configured with a tls authenticator many
720 client connections are expected to not authenticate in this way, so
721 an authenticate fail is not an error.
723 HS/02 Add the Exim version string to the process info. This way exiwhat
724 gives some more detail about the running daemon.
726 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
727 matter for fast-change records such as DNSBLs.
729 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
730 retry record, even if constant. There may be multiple transports with
731 different interface settings and the retry behaviour needs to be kept
734 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
736 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
738 JH/10 Bug 840: fix log_defer_output option of pipe transport
740 JH/11 Bug 830: use same host for all RCPTS of a message, even under
741 hosts_randomize. This matters a lot when combined with mua_wrapper.
743 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
744 ${quote_pgsql:<string>} operator.
746 JH/13 Bug 1708: avoid misaligned access in cached lookup.
748 JH/14 Change header file name for freeradius-client. Relevant if compiling
749 with Radius support; from the Gentoo tree and checked under Fedora.
751 JH/15 Bug 1712: Introduce $prdr_requested flag variable
753 JH/16 Bug 1714: Permit an empty string as expansion result for transport
754 option transport_filter, meaning no filtering.
756 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
758 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
759 defaults to "*" (all hosts). The variable is now available when not built
760 with TLS, default unset, mainly to enable keeping the testsuite sane.
761 If a server certificate is not supplied (via tls_certificate) an error is
762 logged, and clients will find TLS connections fail on startup. Presumably
763 they will retry in-clear.
764 Packagers of Exim are strongly encouraged to create a server certificate
765 at installation time.
767 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
768 with the $config_file variable.
770 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
771 in transport context, after the attempt, and per-recipient. The latter type
772 is per host attempted. The event data is the error message, and the errno
773 information encodes the lookup type (A vs. MX) used for the (first) host,
774 and the trailing two digits of the smtp 4xx response.
776 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
777 to write to mainlog (or rejectlog, paniclog) in the window between file
778 creation and permissions/ownership being changed. Particularly affects
779 installations where exicyclog is run as root, rather than exim user;
780 result is that the running daemon panics and dies.
782 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
784 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
785 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
786 "pri" and "weight". Note that the previous implicit priority given by the
787 list order is no longer honoured.
789 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
792 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
793 by defining SUPPORT_SOCKS.
795 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
796 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
797 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
798 variables are renamed to proxy_{local,external}_{address,port}.
800 JH/27 Move Internationalisation support from Experimental to mainline, enabled
801 for a build by defining SUPPORT_I18N
803 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
804 of the query string, and make ${quote_redis:} do that quoting.
806 JH/29 Move Events support from Experimental to mainline, enabled by default
807 and removable for a build by defining DISABLE_EVENT.
809 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
811 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
812 cached by the daemon.
814 JH/32 Move Redis support from Experimental to mainline, enabled for a build
815 by defining LOOKUP_REDIS. The libhiredis library is required.
817 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
818 keys are given for lookup.
820 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
821 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
822 only supported when built with TLS support. The PolarSSL SHA routines
823 are still used when the TLS library is too old for convenient support.
825 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
826 openssl_options), for security. OpenSSL forces this from version 1.1.0
827 server-side so match that on older versions.
829 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
830 allocation for $value could be released as the expansion processing
831 concluded, but leaving the global pointer active for it.
833 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
834 and to use the domains and local_parts ACL conditions.
836 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
837 incorrectly not doubled on cutthrough transmission, hence seen as a
838 body-termination at the receiving system - resulting in truncated mails.
839 Commonly the sender saw a TCP-level error, and retransmitted the message
840 via the normal store-and-forward channel. This could result in duplicates
841 received - but deduplicating mailstores were liable to retain only the
842 initial truncated version.
844 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
846 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
848 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
849 we're in there, support oversigning also; bug 1309.
851 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
853 HS/04 Add support for keep_environment and add_environment options.
855 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
856 either intentional arithmetic overflow during PRNG, or testing config-
859 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
860 delivery resulted in actual delivery. Cancel cutthrough before DATA
863 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
864 rejects a recipient: pass the reject to the originator.
866 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
867 Many were false-positives and ignorable, but it's worth fixing the
870 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
871 for the new environment-manipulation done at startup. Move the routines
872 from being local to tls.c to being global via the os.c file.
874 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
875 an extract embedded as result-arg for a map, the first arg for extract
876 is unavailable so we cannot tell if this is a numbered or keyed
877 extraction. Accept either.
883 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
886 JH/02 The smtp transport option "multi_domain" is now expanded.
888 JH/03 The smtp transport now requests PRDR by default, if the server offers
891 JH/04 Certificate name checking on server certificates, when exim is a client,
892 is now done by default. The transport option tls_verify_cert_hostnames
893 can be used to disable this per-host. The build option
894 EXPERIMENTAL_CERTNAMES is withdrawn.
896 JH/05 The value of the tls_verify_certificates smtp transport and main options
897 default to the word "system" to access the system default CA bundle.
898 For GnuTLS, only version 3.0.20 or later.
900 JH/06 Verification of the server certificate for a TLS connection is now tried
901 (but not required) by default. The verification status is now logged by
902 default, for both outbound TLS and client-certificate supplying inbound
905 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
908 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
909 Status Notification (bounce) messages are now MIME format per RFC 3464.
910 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
911 under the control of the dsn_advertise_hosts option, and routers may
912 have a dsn_lasthop option.
914 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
915 default, modifiable by a malware= option. The list separator for
916 the options can now be changed in the usual way. Bug 68.
918 JH/10 The smtp_receive_timeout main option is now expanded before use.
920 JH/11 The incoming_interface log option now also enables logging of the
921 local interface on delivery outgoing connections.
923 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
924 if the interface and destination host and port all match.
926 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
929 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
930 Patch from Andrew Lewis.
932 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
933 now supports optional time-restrictions, weighting, and priority
934 modifiers per server. Patch originally by <rommer@active.by>.
936 JH/16 The spamd_address main option now supports a mixed list of local
937 and remote servers. Remote servers can be IPv6 addresses, and
938 specify a port-range.
940 JH/17 Bug 68: The spamd_address main option now supports an optional
941 timeout value per server.
943 JH/18 Bug 1581: Router and transport options headers_add/remove can
944 now have the list separator specified.
946 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
949 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
952 JH/21 Support for the A6 type of dns record is withdrawn.
954 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
955 rather than the verbs used.
957 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
958 from 255 to 1024 chars.
960 JH/24 Verification callouts now attempt to use TLS by default.
962 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
963 are generic router options now. The defaults didn't change.
965 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
966 Original patch from Alexander Shikoff, worked over by JH.
968 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
969 files if ZCAT_COMMAND is not executable.
971 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
973 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
975 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
976 Normally benign, it bites when the pair was led to by a CNAME;
977 modern usage is to not canonicalize the domain to a CNAME target
978 (and we were inconsistent anyway for A-only vs AAAA+A).
980 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
982 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
983 when evaluating $sender_host_dnssec.
985 JH/31 Check the HELO verification lookup for DNSSEC, adding new
986 $sender_helo_dnssec variable.
988 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
990 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
992 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
994 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
995 documented as working, but never had. Support all but $spam_report.
997 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
998 added for tls authenticator.
1000 HS/03 Add perl_taintmode main config option
1006 TL/01 When running the test suite, the README says that variables such as
1007 no_msglog_check are global and can be placed anywhere in a specific
1008 test's script, however it was observed that placement needed to be near
1009 the beginning for it to behave that way. Changed the runtest perl
1010 script to read through the entire script once to detect and set these
1011 variables, reset to the beginning of the script, and then run through
1012 the script parsing/test process like normal.
1014 TL/02 The BSD's have an arc4random API. One of the functions to induce
1015 adding randomness was arc4random_stir(), but it has been removed in
1016 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1017 function when detected.
1019 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1020 cause callback expansion.
1022 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1023 syntax errors in an expansion can be treated as a string instead of
1024 logging or causing an error, due to the internal use of bool_lax
1025 instead of bool when processing it.
1027 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1028 server certificates when making smtp deliveries.
1030 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1032 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1034 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1036 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1037 Merged patch from Sebastian Wiedenroth.
1039 JH/05 Fix results-pipe from transport process. Several recipients, combined
1040 with certificate use, exposed issues where response data items split
1041 over buffer boundaries were not parsed properly. This eventually
1042 resulted in duplicates being sent. This issue only became common enough
1043 to notice due to the introduction of connection certificate information,
1044 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1046 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1047 size buffer was used, resulting in syntax errors when an expansion
1050 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1051 version 3.3.6 or later.
1053 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1054 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1055 both become "event_action", the variables become $event_name, $event_data
1056 and $event_defer_errno. There is a new variable $verify_mode, usable in
1057 routers, transports and related events. The tls:cert event is now also
1058 raised for inbound connections, if the main configuration event_action
1061 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1062 early OCSP support, but no stapling (appears to be less than 1.0.0).
1064 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1065 server certificate names available under the smtp transport option
1066 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1069 JH/10 Time-related extraction expansions from certificates now use the main
1070 option "timezone" setting for output formatting, and are consistent
1071 between OpenSSL and GnuTLS compilations. Bug 1541.
1073 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1074 encoded parameter in the incoming message. Bug 1558.
1076 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1077 include certificate info, eximon was claiming there were spoolfile
1080 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1082 JH/14 Log delivery-related information more consistently, using the sequence
1083 "H=<name> [<ip>]" wherever possible.
1085 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1086 are problematic for Debian distribution, omit them from the release
1089 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1091 JH/16 Fix string representation of time values on 64bit time_t architectures.
1094 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1095 output list separator was used.
1100 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1101 checkers that were complaining about end of non-void function with no
1104 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1105 This was a regression introduced in 4.83 by another bugfix.
1107 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1109 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1110 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1116 TF/01 Correctly close the server side of TLS when forking for delivery.
1118 When a message was received over SMTP with TLS, Exim failed to clear up
1119 the incoming connection properly after forking off the child process to
1120 deliver the message. In some situations the subsequent outgoing
1121 delivery connection happened to have the same fd number as the incoming
1122 connection previously had. Exim would try to use TLS and fail, logging
1123 a "Bad file descriptor" error.
1125 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1126 utilities have not been installed.
1128 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1129 temporary space as the ACL may create new global variables.
1131 TL/01 LDAP support uses per connection or global context settings, depending
1132 upon the detected version of the libraries at build time.
1134 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1135 to extract and use the src ip:port in logging and expansions as if it
1136 were a direct connection from the outside internet. PPv2 support was
1137 updated based on HAProxy spec change in May 2014.
1139 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1141 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1142 Properly escape header and check for NULL return.
1144 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1147 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1149 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1150 characters in header names, implemented as a verify condition.
1151 Contributed by Michael Fischer v. Mollard.
1153 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1154 results permerror and temperror. Previous values are deprecated but
1155 still accepted. In a future release, err_perm and err_temp will be
1156 completely removed, which will be a backward incompatibility if the
1157 ACL tests for either of these two old results. Patch contributed by
1158 user bes-internal on the mailing list.
1160 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1162 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1163 selectors, in both main and reject logs.
1165 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1168 JH/07 Add malware type "sock" for talking to simple daemon.
1170 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1172 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1173 routers/transports under cutthrough routing.
1175 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1176 numbers. Touch up "bool" conditional to keep the same definition.
1178 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1180 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1182 JH/12 Expand items in router/transport headers_add or headers_remove lists
1183 individually rather than the list as a whole. Bug 1452.
1185 Required for reasonable handling of multiple headers_ options when
1186 they may be empty; requires that headers_remove items with embedded
1187 colons must have them doubled (or the list-separator changed).
1189 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1190 view the policy declared in the DMARC record. Currently, $dmarc_status
1191 is a combined value of both the record presence and the result of the
1194 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1196 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1197 dnslookup router and the smtp transport (applying to the forward
1200 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1201 of ldap servers used for a specific lookup. Patch provided by Heiko
1204 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1205 New variable $lookup_dnssec_authenticated for observability.
1207 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1208 Patch submitted by Lars Timman.
1210 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1212 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1213 Requires trusted mode and valid format message id, aborts otherwise.
1214 Patch contributed by Heiko Schlichting.
1216 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1217 certextract with support for various fields. Bug 1358.
1219 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1220 is requested by default, modifiable by smtp transport option
1223 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1224 operate on certificate variables to give certificate fingerprints
1225 Also new ${sha256:cert_variable}.
1227 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1229 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1230 Christian Aistleitner.
1232 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1234 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1235 file. Patch from Wolfgang Breyha.
1237 JH/25 Expand the coverage of the delivery $host and $host_address to
1238 client authenticators run in verify callout. Bug 1476.
1240 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1241 align with daemon_smtp_ports. Bug 72.
1243 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1244 support and error reporting did not work properly.
1246 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1247 and is readable. Patch from Andrew Colin Kissa.
1249 TL/14 Enhance documentation of ${run expansion and how it parses the
1250 commandline after expansion, particularly in the case when an
1251 unquoted variable expansion results in an empty value.
1253 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1255 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1256 by renaming away from it. Addresses GH issue 15, reported by
1259 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1261 TL/15 SECURITY: prevent double expansion in math comparison functions
1262 (can expand unsanitized data). Not remotely exploitable.
1269 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1271 PP/02 Make -n do something, by making it not do something.
1272 When combined with -bP, the name of an option is not output.
1274 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1277 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1278 $sender_host_name and config options to manage this, and basic check
1281 PP/05 DSCP support for outbound connections and control modifier for inbound.
1283 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1284 (Only plugin which currently uses this is kerberos4, which nobody should
1285 be using, but we should make it available and other future plugins might
1286 conceivably use it, even though it would break NAT; stuff *should* be
1287 using channel bindings instead).
1289 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1290 name; added for Sendmail compatibility; requires admin caller.
1291 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1292 just ignore it); requires trusted caller.
1293 Also parse but ignore: -Ac -Am -X<logfile>
1296 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1298 TL/02 Add +smtp_confirmation as a default logging option.
1300 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1301 Patch by Magnus Holmgren from 2007-02-20.
1303 TL/04 Bugzilla 1281 - Spec typo.
1304 Bugzilla 1283 - Spec typo.
1305 Bugzilla 1290 - Spec grammar fixes.
1307 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1309 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1311 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1312 mailing list by by Dmitry Isaikin.
1314 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1316 JH/02 Support "G" suffix to numbers in ${if comparisons.
1318 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1320 NM/01 Bugzilla 1197 - Spec typo
1321 Bugzilla 1196 - Spec examples corrections
1323 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1325 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1326 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1329 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1330 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1332 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1333 "acl {{name}{arg}...}", and optional args on acl condition
1336 JH/05 Permit multiple router/transport headers_add/remove lines.
1338 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1340 JH/07 Avoid using a waiting database for a single-message-only transport.
1341 Performance patch from Paul Fisher. Bugzilla 1262.
1343 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1346 JH/09 Add $headers_added variable, with content from use of ACL modifier
1347 add_header (but not yet added to the message). Bugzilla 199.
1349 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1350 Pulled from Bugzilla 817 by Wolfgang Breyha.
1352 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1354 (nb: this is the same fix as in Exim 4.80.1)
1356 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1359 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1362 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1364 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1365 advertise SMTP AUTH mechanism to us, instead of a generic
1366 protocol violation error. Also, make Exim more robust to bad
1367 data from the Dovecot auth socket.
1369 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1371 When a queue runner is handling a message, Exim first routes the
1372 recipient addresses, during which it prunes them based on the retry
1373 hints database. After that it attempts to deliver the message to
1374 any remaining recipients. It then updates the hints database using
1377 So if a recipient address works intermittently, it can get repeatedly
1378 deferred at routing time. The retry hints record remains fresh so the
1379 address never reaches the final cutoff time.
1381 This is a fairly common occurrence when a user is bumping up against
1382 their storage quota. Exim had some logic in its local delivery code
1383 to deal with this. However it did not apply to per-recipient defers
1384 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1386 This change adds a proper retry rule check during routing so that the
1387 final cutoff time is checked against the message's age. We only do
1388 this check if there is an address retry record and there is not a
1389 domain retry record; this implies that previous attempts to handle
1390 the address had the retry_use_local_parts option turned on. We use
1391 this as an approximation for the destination being like a local
1392 delivery, as in LMTP.
1394 I suspect this new check makes the old local delivery cutoff check
1395 redundant, but I have not verified this so I left the code in place.
1397 TF/02 Correct gecos expansion when From: is a prefix of the username.
1399 Test 0254 submits a message to Exim with the header
1403 When I ran the test suite under the user fanf2, Exim expanded
1404 the header to contain my full name, whereas it should have added
1405 a Resent-Sender: header. It erroneously treats any prefix of the
1406 username as equal to the username.
1408 This change corrects that bug.
1410 GF/01 DCC debug and logging tidyup
1411 Error conditions log to paniclog rather than rejectlog.
1412 Debug lines prefixed by "DCC: " to remove any ambiguity.
1414 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1416 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1417 Bug spotted by Jeremy Harris; was flawed since initial commit.
1418 Would have resulted in OCSP responses post-SNI triggering an Exim
1419 NULL dereference and crash.
1421 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1423 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1424 Bug detection, analysis and fix by Samuel Thibault.
1425 Bugzilla 1331, Debian bug #698092.
1427 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1429 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1430 Server implementation by Todd Lyons, client by JH.
1431 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1432 config variable "prdr_enable" controls whether the server
1433 advertises the facility. If the client requests PRDR a new
1434 acl_data_smtp_prdr ACL is called once for each recipient, after
1435 the body content is received and before the acl_smtp_data ACL.
1436 The client is controlled by both of: a hosts_try_prdr option
1437 on the smtp transport, and the server advertisement.
1438 Default client logging of deliveries and rejections involving
1439 PRDR are flagged with the string "PRDR".
1441 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1442 fclose(). Diagnosis by Todd Lyons.
1444 PP/17 Update configure.default to handle IPv6 localhost better.
1445 Patch by Alain Williams (plus minor tweaks).
1448 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1449 This is now consistent with GnuTLS, and is now documented: the
1450 previous undocumented portable approach to treating the option as
1451 unset was to force an expansion failure. That still works, and
1452 an empty string is now equivalent.
1454 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1455 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1456 not performing validation itself.
1458 PP/20 Added force_command boolean option to pipe transport.
1459 Patch from Nick Koston, of cPanel Inc.
1461 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1464 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1466 PP/21 Fix eximon continuous updating with timestamped log-files.
1467 Broken in a format-string cleanup in 4.80, missed when I repaired the
1468 other false fix of the same issue.
1469 Report and fix from Heiko Schlichting.
1472 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1473 Report from Prashanth Katuri.
1475 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1476 It's SecureTransport, so affects any MacOS clients which use the
1477 system-integrated TLS libraries, including email clients.
1479 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1480 using a MIME ACL for non-SMTP local injection.
1481 Report and assistance in diagnosis by Warren Baker.
1483 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1485 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1487 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1488 last id that failed so it may be referenced in subsequent ACL's.
1490 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1493 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1494 ldap library initialization, allowing self-signed CA's to be
1495 used. Also properly sets require_cert option later in code by
1496 using NULL (global ldap config) instead of ldap handle (per
1497 session). Bug diagnosis and testing by alxgomz.
1499 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1500 the src/util/ subdirectory.
1502 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1503 renamed to Transport Post Delivery Action by Jeremy Harris, as
1506 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1507 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1508 redis_servers = needs to be configured which will be used by the redis
1509 lookup. Patch from Warren Baker, of The Packet Hub.
1511 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1513 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1514 hostname or reverse DNS when processing a host list. Used suggestions
1515 from multiple comments on this bug.
1517 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1519 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1520 Missed a few lines, added it to make the runtest require no keyboard
1523 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1524 contains upper case chars. Make router use caseful_local_part.
1526 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1527 support when GnuTLS has been built with p11-kit.
1533 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1535 This, or similar/improved, will also be change PP/11 of 4.82.
1541 PP/01 Handle short writes when writing local log-files.
1542 In practice, only affects FreeBSD (8 onwards).
1543 Bugzilla 1053, with thanks to Dmitry Isaikin.
1545 NM/01 Bugzilla 949 - Documentation tweak
1547 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1550 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1552 PP/02 Implemented gsasl authenticator.
1554 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1556 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1557 `pkg-config foo` for cflags/libs.
1559 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1560 with rest of GSASL and with heimdal_gssapi.
1562 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1563 `pkg-config foo` for cflags/libs for the TLS implementation.
1565 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1566 properties get this fed in as external SSF. A number of robustness
1567 and debugging improvements to the cyrus_sasl authenticator.
1569 PP/08 cyrus_sasl server now expands the server_realm option.
1571 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1572 Patch by Jeremy Harris.
1574 PP/10 Added dbmjz lookup type.
1576 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1578 PP/12 MAIL args handles TAB as well as SP, for better interop with
1579 non-compliant senders.
1580 Analysis and variant patch by Todd Lyons.
1582 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1583 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1584 Patch from Dirk Mueller <dmueller@suse.com>
1586 PP/13 tls_peerdn now print-escaped for spool files.
1587 Observed some $tls_peerdn in wild which contained \n, which resulted
1588 in spool file corruption.
1590 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1591 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1592 or write after TLS renegotiation, which otherwise led to messages
1595 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1596 as a tracking header (ie: a signed header comes before the signature).
1597 Patch from Wolfgang Breyha.
1599 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1600 comma-sep list; embedded commas doubled.
1602 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1604 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1606 Report and patch from Dmitry Banschikov.
1608 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1609 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1610 failures. We appear to now support TLS1.1+ with Exim.
1612 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1613 lets Exim select keys and certificates based upon TLS SNI from client.
1614 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1615 before an outbound SMTP session. New log_selector, +tls_sni.
1617 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1618 NULL dereference. Report and patch from Alun Jones.
1620 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1621 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1622 Not seeing resolver debug output on NetBSD, but suspect this is a
1623 resolver implementation change.
1625 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1626 Left warnings. Added "eximon gdb" invocation mode.
1628 PP/21 Defaulting "accept_8bitmime" to true, not false.
1630 PP/22 Added -bw for inetd wait mode support.
1632 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1633 locate the relevant includes and libraries. Made this the default.
1635 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1636 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1638 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1639 This may cause build issues on older platforms.
1641 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1642 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1643 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1644 Added SNI support via GnuTLS too.
1645 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1647 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1649 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1650 Applied second patch from Janne, implementing suggestion to default
1651 multiple-strings-in-record handling to match SPF spec.
1653 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1655 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1656 read-only, out of scope).
1657 Patch from Wolfgang Breyha, report from Stuart Northfield.
1659 PP/29 Fix three issues highlighted by clang analyser static analysis.
1660 Only crash-plausible issue would require the Cambridge-specific
1661 iplookup router and a misconfiguration.
1662 Report from Marcin Mirosław.
1664 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1666 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1667 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1668 As part of this, removing so much warning spew let me fix some minor
1669 real issues in debug logging.
1671 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1672 assignment on my part. Fixed.
1674 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1675 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1676 Janne Snabb (who went above and beyond: thank you).
1678 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1679 string otherwise requires a connection and a bunch more work and it's
1680 relatively easy to get wrong. Should also expose TLS library linkage
1683 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1684 64-bit ${eval} (JH/03).
1686 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1687 GNU libc to support some of the 64-bit stuff, should not lead to
1688 conflicts. Defined before os.h is pulled in, so if a given platform
1689 needs to override this, it can.
1691 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1692 protection layer was required, which is not implemented.
1693 Bugzilla 1254, patch from Wolfgang Breyha.
1695 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1696 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1697 tls_dhparam take prime identifiers. Also unbreak combination of
1698 OpenSSL+DH_params+TLSSNI.
1700 PP/39 Disable SSLv2 by default in OpenSSL support.
1706 PP/01 Solaris build fix for Oracle's LDAP libraries.
1707 Bugzilla 1109, patch from Stephen Usher.
1709 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1711 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1714 TF/02 Fix a couple more cases where we did not log the error message
1715 when unlink() failed. See also change 4.74-TF/03.
1717 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1718 lock up or crash if it happened to be inside a call to libc when it
1719 got a SIGUSR1 from exiwhat.
1721 The SIGUSR1 handler appends the current process status to the process
1722 log which is later printed by exiwhat. It used to use the general
1723 purpose logging code to do this, but several functions it calls are
1724 not safe for signals.
1726 The new output code in the SIGUSR1 handler is specific to the process
1727 log, and simple enough that it's easy to inspect for signal safety.
1728 Removing some special cases also simplifies the general logging code.
1729 Removing the spurious timestamps from the process log simplifies
1732 TF/04 Improved ratelimit ACL condition.
1734 The /noupdate option has been deprecated in favour of /readonly which
1735 has clearer semantics. The /leaky, /strict, and /readonly update modes
1736 are mutually exclusive. The update mode is no longer included in the
1737 database key; it just determines when the database is updated. (This
1738 means that when you upgrade Exim will forget old rate measurements.)
1740 Exim now checks that the per_* options are used with an update mode that
1741 makes sense for the current ACL. For example, when Exim is processing a
1742 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1743 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1744 must specify per_mail/readonly. If you omit the update mode it defaults to
1745 /leaky where that makes sense (as before) or /readonly where required.
1747 The /noupdate option is now undocumented but still supported for
1748 backwards compatibility. It is equivalent to /readonly except that in
1749 ACLs where /readonly is required you may specify /leaky/noupdate or
1750 /strict/noupdate which are treated the same as /readonly.
1752 A useful new feature is the /count= option. This is a generalization
1753 of the per_byte option, so that you can measure the throughput of other
1754 aggregate values. For example, the per_byte option is now equivalent
1755 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1757 The per_rcpt option has been generalized using the /count= mechanism
1758 (though it's more complicated than the per_byte equivalence). When it is
1759 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1760 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1761 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1762 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1763 Note that using per_rcpt with a non-readonly update mode in more than
1764 one ACL will cause the recipients to be double-counted. (The per_mail
1765 and per_byte options don't have this problem.)
1767 The handling of very low rates has changed slightly. If the computed rate
1768 is less than the event's count (usually one) then this event is the first
1769 after a long gap. In this case the rate is set to the same as this event's
1770 count, so that the first message of a spam run is counted properly.
1772 The major new feature is a mechanism for counting the rate of unique
1773 events. The new per_addr option counts the number of different
1774 recipients that someone has sent messages to in the last time period. It
1775 behaves like per_rcpt if all the recipient addresses are different, but
1776 duplicate recipient addresses do not increase the measured rate. Like
1777 the /count= option this is a general mechanism, so the per_addr option
1778 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1779 example, measure the rate that a client uses different sender addresses
1780 with the options per_mail/unique=$sender_address. There are further
1781 details in the main documentation.
1783 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1785 TF/06 Removed a few PCRE remnants.
1787 TF/07 Automatically extract Exim's version number from tags in the git
1788 repository when doing development or release builds.
1790 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1791 Bugzilla 879. Patch from Paul Fisher.
1793 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1794 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1797 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1799 PP/05 Variable $av_failed, true if the AV scanner deferred.
1800 Bugzilla 1078. Patch from John Horne.
1802 PP/06 Stop make process more reliably on build failure.
1803 Bugzilla 1087. Patch from Heiko Schlittermann.
1805 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1806 Bugzilla 1089. Patch from Heiko Schlittermann.
1808 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1809 Bugzilla 1131. Patch from Holger Weiß.
1811 PP/09 Handle IPv6 addresses with SPF.
1812 Bugzilla 860. Patch from Wolfgang Breyha.
1814 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1816 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1819 PP/11 match_* no longer expand right-hand-side by default.
1820 New compile-time build option, EXPAND_LISTMATCH_RHS.
1821 New expansion conditions, "inlist", "inlisti".
1823 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1825 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1827 PP/14 fix log_write() format string regression from TF/03.
1828 Bugzilla 1152. Patch from Dmitry Isaikin.
1834 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1836 PP/02 Harmonised TLS library version reporting; only show if debugging.
1837 Layout now matches that introduced for other libraries in 4.74 PP/03.
1839 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1841 PP/04 New "dns_use_edns0" global option.
1843 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1846 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1847 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1849 TK/01 Updated PolarSSL code to 0.14.2.
1850 Bugzilla 1097. Patch from Andreas Metzler.
1852 PP/07 Catch divide-by-zero in ${eval:...}.
1853 Fixes bugzilla 1102.
1855 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1858 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1859 format-string attack -- SECURITY: remote arbitrary code execution.
1861 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1862 time unintentionally subject to list matching rules, letting the header
1863 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1864 arbitrary string expansion). This allowed for information disclosure.
1866 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1867 INT_MIN/-1 -- value coerced to INT_MAX.
1873 NM/01 Workaround for PCRE version dependency in version reporting
1876 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1877 This fixes portability to compilers other than gcc, notably
1878 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1880 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1881 makefiles for portability to HP-UX and POSIX correctness.
1883 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1884 Also via indented variable definition in the Makefile.
1885 (Debugging by Oliver Heesakkers).
1887 PP/02 Restore caching of spamd results with expanded spamd_address.
1888 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1890 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1891 Improves build reliability. Fix from: Frank Elsner
1893 NM/02 Fix wide character breakage in the rfc2047 coding
1894 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1896 NM/03 Allow underscore in dnslist lookups
1897 Fixes bug 1026. Patch from Graeme Fowler
1899 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1900 Code patches from Adam Ciarcinski of NetBSD.
1902 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1905 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1906 is logged, to avoid truncation. Patch from John Horne.
1908 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1909 Patch from Jakob Hirsch.
1911 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1912 SQL string expansion failure details.
1913 Patch from Andrey Oktyabrski.
1915 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1916 Patch from Simon Arlott.
1918 PP/09 New lookups functionality failed to compile on old gcc which rejects
1919 extern declarations in function scope.
1920 Patch from Oliver Fleischmann
1922 PP/10 Use sig_atomic_t for flags set from signal handlers.
1923 Check getgroups() return and improve debugging.
1924 Fixed developed for diagnosis in bug 927 (which turned out to be
1927 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1928 Patch from Mark Zealey.
1930 PP/12 Bugzilla 1056: Improved spamd server selection.
1931 Patch from Mark Zealey.
1933 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1934 Based on patch from Heiko Schlittermann.
1936 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1937 Patch from Uwe Doering, sign-off by Michael Haardt.
1939 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1940 variable declaration deep within a block. Bug and patch from
1943 PP/15 lookups-Makefile IRIX compatibility coercion.
1945 PP/16 Make DISABLE_DKIM build knob functional.
1947 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1948 Patch by Simon Arlott
1950 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1951 variable argument macros. Our copy now differs from upstream.
1957 TF/01 Failure to get a lock on a hints database can have serious
1958 consequences so log it to the panic log.
1960 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1961 controlled using the smtp_confirmation log selector.
1963 TF/03 Include the error message when we fail to unlink a spool file.
1965 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1966 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1967 for maintaining out-of-tree patches for some time.
1969 PP/01 Bugzilla 139: Documentation and portability issues.
1970 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1971 Handle per-OS dynamic-module compilation flags.
1973 PP/02 Let /dev/null have normal permissions.
1974 The 4.73 fixes were a little too stringent and complained about the
1975 permissions on /dev/null. Exempt it from some checks.
1976 Reported by Andreas M. Kirchwitz.
1978 PP/03 Report version information for many libraries, including
1979 Exim version information for dynamically loaded libraries. Created
1980 version.h, now support a version extension string for distributors
1981 who patch heavily. Dynamic module ABI change.
1983 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1984 privilege escalation vulnerability whereby the Exim run-time user
1985 can cause root to append content of the attacker's choosing to
1988 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1991 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1992 If dropping privileges for untrusted macros, we disabled normal logging
1993 on the basis that it would fail; for the Exim run-time user, this is not
1994 the case, and it resulted in successful deliveries going unlogged.
1995 Fixed. Reported by Andreas Metzler.
2001 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2002 only prepend for the Resent-* case. Fixes regression introduced in
2003 Exim 4.70 by NM/22 for Bugzilla 607.
2005 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2006 increasing numbers of administrators be bitten by this.
2008 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2010 PP/03 Bugzilla 994: added openssl_options main configuration option.
2012 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2014 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2016 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2018 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2019 without a peer certificate, leading to a segfault because of an
2020 assumption that peers always have certificates. Be a little more
2021 paranoid. Problem reported by Martin Tscholak.
2023 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2024 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2025 NB: ClamAV planning to remove STREAM in "middle of 2010".
2026 CL also introduces -bmalware, various -d+acl logging additions and
2027 more caution in buffer sizes.
2029 PP/09 Implemented reverse_ip expansion operator.
2031 PP/10 Bugzilla 937: provide a "debug" ACL control.
2033 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2035 PP/12 Bugzilla 973: Implement --version.
2037 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2039 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2041 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2043 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2044 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2045 ignore trailing whitespace.
2047 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2049 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2052 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2053 Notification from Dr Andrew Aitchison.
2055 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2056 ExtendedDetectionInfo response format.
2057 Notification from John Horne.
2059 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2062 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2063 XSL and documented dependency on system catalogs, with examples of how
2066 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2069 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2070 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2071 configuration file which is writeable by the Exim user or group.
2073 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2074 of configuration files to cover files specified with the -C option if
2075 they are going to be used with root privileges, not just the default
2078 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2079 option (effectively making it always true).
2081 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2082 files to be used while preserving root privileges.
2084 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2085 that rogue child processes cannot use them.
2087 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2088 run-time user, instead of root.
2090 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2091 Exim run-time user without dropping privileges.
2093 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2094 result string, instead of calling string_vformat() twice with the same
2097 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2098 for other users. Others should always drop root privileges if they use
2099 -C on the command line, even for a whitelisted configure file.
2101 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2103 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2109 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2110 $data_path, and $header_path variables; fixed documentation bugs and
2113 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2114 exipick to access non-standard spools, including the "frozen" queue
2117 NM/01 Bugzilla 965: Support mysql stored procedures.
2118 Patch from Alain Williams
2120 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2122 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2123 Patch from Andreas Metzler
2125 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2126 Patch from Kirill Miazine
2128 NM/05 Bugzilla 671: Added umask to procmail example.
2130 JJ/03 installed exipick 20100323.0, fixing doc bug
2132 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2133 directory. Notification and patch from Dan Rosenberg.
2135 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2137 TK/02 Improve log output when DKIM signing operation fails.
2139 MH/01 Treat the transport option dkim_domain as a colon separated
2140 list, not as a single string, and sign the message with each element,
2141 omitting multiple occurences of the same signer.
2143 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2144 Bugzilla 985, 986. Patch by Simon Arlott
2146 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2147 Patch by Simon Arlott
2149 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2150 MBX locking. Notification from Dan Rosenberg.
2156 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2158 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2160 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2162 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2164 NM/04 Bugzilla 915: Flag broken perl installation during build.
2170 TK/01 Added patch by Johannes Berg that expands the main option
2171 "spamd_address" if it starts with a dollar sign.
2173 TK/02 Write list of recipients to X-Envelope-Sender header when building
2174 the mbox-format spool file for content scanning (suggested by Jakob
2177 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2178 (http://www.dcc-servers.net/) support via dccifd. Activated by
2179 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2181 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2182 by Mark Daniel Reidel <mr@df.eu>.
2184 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2185 When building exim an external PCRE library is now needed -
2186 PCRE is a system library on the majority of modern systems.
2187 See entry on PCRE_LIBS in EDITME file.
2189 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2190 conversation. Added nologin parameter to request.
2191 Patch contributed by Kirill Miazine.
2193 TF/01 Do not log submission mode rewrites if they do not change the address.
2195 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2197 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2198 log files in place. Contributed by Roberto Lima.
2200 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2202 TF/03 Bugzilla 615: When checking the local_parts router precondition
2203 after a local_part_suffix or local_part_prefix option, Exim now
2204 does not use the address's named list lookup cache, since this
2205 contains cached lookups for the whole local part.
2207 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2208 Robert Millan. Documentation is in experimental-spec.txt.
2210 TF/04 Bugzilla 668: Fix parallel build (make -j).
2212 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2214 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2215 Patch provided by Jan Srzednicki.
2217 TF/05 Leading white space used to be stripped from $spam_report which
2218 wrecked the formatting. Now it is preserved.
2220 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2221 that they are available at delivery time.
2223 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2225 TF/08 TLS error reporting now respects the incoming_interface and
2226 incoming_port log selectors.
2228 TF/09 Produce a more useful error message if an SMTP transport's hosts
2229 setting expands to an empty string.
2231 NM/06 Bugzilla 744: EXPN did not work under TLS.
2232 Patch provided by Phil Pennock.
2234 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2235 Patch provided by Richard Godbee.
2237 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2238 acl_smtp_notquit, added index entry.
2240 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2241 Patch provided by Eugene Bujak.
2243 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2244 accept(). Patch provided by Maxim Dounin.
2246 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2247 Patch provided by Phil Pennock.
2249 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2251 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2252 Patch provided by Brad "anomie" Jorsch.
2254 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2255 Patch provided by Dean Brooks.
2257 TK/05 Add native DKIM support (does not depend on external libraries).
2259 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2260 Patch provided by Graeme Fowler.
2262 NM/16 Bugzilla 851: Documentation example syntax fix.
2264 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2266 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2269 NM/19 Bugzilla 745: TLS version reporting.
2270 Patch provided by Phil Pennock.
2272 NM/20 Bugzilla 167: bool: condition support.
2273 Patch provided by Phil Pennock.
2275 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2276 clients. Patch provided by Phil Pennock.
2278 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2279 Patch provided by Brad "anomie" Jorsch.
2281 NM/23 Bugzilla 687: Fix misparses in eximstats.
2282 Patch provided by Heiko Schlittermann.
2284 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2285 Patch provided by Heiko Schlittermann.
2287 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2288 plus update to original patch.
2290 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2292 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2293 Patch provided by David Brownlee.
2295 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2297 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2299 NM/30 Bugzilla 888: TLS documentation bugfixes.
2301 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2303 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2304 Unlike the original bugzilla I have changed all shell scripts in src tree.
2306 NM/33 Bugzilla 898: Transport filter timeout fix.
2307 Patch by Todd Rinaldo.
2309 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2310 Patch by Serge Demonchaux.
2312 NM/35 Bugzilla 39: Base64 decode bug fixes.
2313 Patch by Jakob Hirsch.
2315 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2317 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2319 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2321 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2327 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2328 ALT-N's libdkim that I have put here:
2329 http://duncanthrax.net/exim-experimental/
2331 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2332 were called 'true' and it seems that C99 defines that as a reserved
2333 keyword to be used with 'bool' variable types. That means you could
2334 not include C99-style headers which use bools without triggering
2335 build errors in sieve.c.
2337 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2338 as mailq or other aliases. Changed the --help handling significantly
2339 to do whats expected. exim_usage() emits usage/help information.
2341 SC/01 Added the -bylocaldomain option to eximstats.
2343 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2345 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2347 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2353 PH/01 Another patch from the Sieve maintainer.
2355 PH/02 When an IPv6 address is converted to a string for single-key lookup
2356 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2357 dots are used instead of colons so that keys in lsearch files need not
2358 contain colons. This was done some time before quoting was made available
2359 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2360 (notated using the quote facility) so as to distinguish them from IPv4
2361 keys. This meant that lookups for IP addresses in host lists did not work
2362 for iplsearch lookups.
2364 This has been fixed by arranging for IPv6 addresses to be expressed with
2365 colons if the lookup type is iplsearch. This is not incompatible, because
2366 previously such lookups could never work.
2368 The situation is now rather anomalous, since one *can* have colons in
2369 ordinary lsearch keys. However, making the change in all cases is
2370 incompatible and would probably break a number of configurations.
2372 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2375 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2376 conversion specification without a maximum field width, thereby enabling
2377 a rogue spamd server to cause a buffer overflow. While nobody in their
2378 right mind would setup Exim to query an untrusted spamd server, an
2379 attacker that gains access to a server running spamd could potentially
2380 exploit this vulnerability to run arbitrary code as the Exim user.
2382 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2383 $primary_hostname instead of what libspf2 thinks the hosts name is.
2385 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2386 a directory entry by the name of the lookup key. Previously, if a
2387 symlink pointed to a non-existing file or a file in a directory that
2388 Exim lacked permissions to read, a lookup for a key matching that
2389 symlink would fail. Now it is enough that a matching directory entry
2390 exists, symlink or not. (Bugzilla 503.)
2392 PH/03 The body_linecount and body_zerocount variables are now exported in the
2395 PH/04 Added the $dnslist_matched variable.
2397 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2398 This means they are set thereafter only if the connection becomes
2401 PH/06 Added the client_condition to authenticators so that some can be skipped
2402 by clients under certain conditions.
2404 PH/07 The error message for a badly-placed control=no_multiline_responses left
2405 "_responses" off the end of the name.
2407 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2409 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2410 (without spaces) instead of just copying the configuration text.
2412 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2414 PH/11 Added $max_received_linelength.
2416 PH/12 Added +ignore_defer and +include_defer to host lists.
2418 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2419 way in which PCRE > 7.0 is built.
2421 PH/14 Implemented queue_only_load_latch.
2423 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2424 MAIL command. The effect was to mangle the value on 64-bit systems.
2426 PH/16 Another patch from the Sieve maintainer.
2428 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2430 PH/18 If a system quota error occurred while trying to create the file for
2431 a maildir delivery, the message "Mailbox is full" was not appended to the
2432 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2433 only to a quota excession during the actual writing of the file.
2435 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2436 characters?) which causes problems in log lines. The DN values are now
2437 passed through string_printing() before being added to log lines.
2439 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2440 and InterBase are left for another time.)
2442 PH/21 Added message_body_newlines option.
2444 PH/22 Guard against possible overflow in moan_check_errorcopy().
2446 PH/23 POSIX allows open() to be a macro; guard against that.
2448 PH/24 If the recipient of an error message contained an @ in the local part
2449 (suitably quoted, of course), incorrect values were put in $domain and
2450 $local_part during the evaluation of errors_copy.
2456 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2457 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2460 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2461 issue a MAIL command.
2463 PH/02 In an ACL statement such as
2465 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2467 if a client was not listed at all, or was listed with a value other than
2468 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2469 the condition was not true (as it should be), so access was not denied.
2470 The bug was that the ! inversion was incorrectly passed on to the second
2471 item. This has been fixed.
2473 PH/03 Added additional dnslists conditions == and =& which are different from
2474 = and & when the dns lookup returns more than one IP address.
2476 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2477 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2479 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2480 FSYNC, which compiles an option called disable_fsync that allows for
2481 bypassing fsync(). The documentation is heavily laced with warnings.
2483 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2485 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2486 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2487 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2488 including adding "make clean"; (3) Added -fPIC when compiling the test
2489 dynamically loaded module, to get rid of a warning.
2491 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2492 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2493 = 0s. The bug is otherwise harmless.
2495 PH/07 There was a bug in the dovecot authenticator such that the value of
2496 $auth1 could be overwritten, and so not correctly preserved, after a
2497 successful authentication. This usually meant that the value preserved by
2498 the server_setid option was incorrect.
2500 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2502 PH/09 Installed PCRE release 7.0.
2504 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2505 run for batched SMTP input. It is now run at the start of every message
2506 in the batch. While fixing this I discovered that the process information
2507 (output by running exiwhat) was not always getting set for -bs and -bS
2508 input. This is fixed, and it now also says "batched" for BSMTP.
2510 PH/11 Added control=no_pipelining.
2512 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2513 patch, slightly modified), and move the expansion of helo_data till after
2514 the connection is made in the smtp transport (so it can use these
2517 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2519 PH/14 Added log_selector = +pid.
2521 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2523 PH/16 Add ${if forany and ${if forall.
2525 PH/17 Added dsn_from option to vary the From: line in DSNs.
2527 PH/18 Flush SMTP output before performing a callout, unless control =
2528 no_callout_flush is set.
2530 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2531 was true (the default) a successful delivery failed to delete the retry
2532 item, thus causing premature timeout of the address. The bug is now
2535 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2537 PH/21 Long custom messages for fakedefer and fakereject are now split up
2538 into multiline responses in the same way that messages for "deny" and
2539 other ACL rejections are.
2541 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2542 with slight modification.
2544 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2545 draft, changing the syntax and factoring some duplicate code".
2547 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2548 for deliveries of the second and subsequent messages over the same SMTP
2551 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2552 ${reduce, with only minor "tidies".
2554 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2556 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2557 expansion side effects.
2559 PH/27 When a message times out after an over-quota error from an Exim-imposed
2560 quota, the bounce message says "mailbox is full". This message was not
2561 being given when it was a system quota that was exceeded. It now should
2564 MH/03 Made $recipients available in local_scan(). local_scan() already has
2565 better access to the recipient list through recipients_list[], but
2566 $recipients can be useful in postmaster-provided expansion strings.
2568 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2569 in the case of a MAIL command with additional options following the
2570 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2571 were accidentally chopped off.
2573 PH/29 SMTP synchronization checks are implemented when a command is read -
2574 there is a check that no more input is waiting when there shouldn't be
2575 any. However, for some commands, a delay in an ACL can mean that it is
2576 some time before the response is written. In this time, more input might
2577 arrive, invalidly. So now there are extra checks after an ACL has run for
2578 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2579 pipelining has not been advertised.
2581 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2583 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2584 returned in the SMTP response when smtp_return_error_details was false.
2585 This has been fixed.
2587 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2588 instead of the C I/O that was originally supplied, because problems were
2589 reported on Solaris.
2591 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2592 Exim which did not show up earlier: it was assuming that a call to
2593 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2594 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2595 was a macro that became an assignment, so it seemed to work. This has
2596 changed to a proper function call with a void return, hence the compile
2597 error. Exim's code has been fixed.
2599 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2602 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2604 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2606 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2607 because it turns out to be too big or there is a timeout) there is no
2608 "Completed" line in the log. When some messages of this type were
2609 selected by exigrep, they were listed as "not completed". Others were
2610 picked up by some special patterns. I have improved the selection
2611 criteria to be more general.
2613 PH/38 The host_find_failed option in the manualroute router can now be set
2614 to "ignore", to completely ignore a host whose IP address cannot be
2615 found. If all hosts are ignored, the behaviour is controlled by the new
2616 host_all_ignored option.
2618 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2619 homing or because of multiple MX records with /mx) generated more than
2620 one IP address, and the following item turned out to be the local host,
2621 all the secondary addresses of the first item were incorrectly removed
2622 from the list, along with the local host and any following hosts (which
2623 is what is supposed to happen).
2625 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2626 whoever called Exim into the -H file. In the case of the daemon it was
2627 behaving confusingly. When first started, it used values for whoever
2628 started the daemon, but after a SIGHUP it used the Exim user (because it
2629 calls itself on a restart). I have changed the code so that it now always
2632 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2633 message are rejected with the same error (e.g. no authentication or bad
2634 sender address), and a DATA command is nevertheless sent (as can happen
2635 with PIPELINING or a stupid MUA), the error message that was given to the
2636 RCPT commands is included in the rejection of the DATA command. This is
2637 intended to be helpful for MUAs that show only the final error to their
2640 PH/42 Another patch from the Sieve maintainer.
2642 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2643 Eximstats - Fixed some broken HTML links and added missing column headers
2645 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2646 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2648 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2650 PH/43 Yet another patch from the Sieve maintainer.
2652 PH/44 I found a way to check for a TCP/IP connection going away before sending
2653 the response to the final '.' that terminates a message, but only in the
2654 case where the client has not sent further data following the '.'
2655 (unfortunately, this is allowed). However, in many cases there won't be
2656 any further data because there won't be any more messages to send. A call
2657 to select() can be used: if it shows that the input is "ready", there is
2658 either input waiting, or the socket has been closed. An attempt to read
2659 the next input character can distinguish the two cases. Previously, Exim
2660 would have sent an OK response which the client would never have see.
2661 This could lead to message repetition. This fix should cure that, at
2662 least in a lot of common cases.
2664 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2665 advertised in response to EHLO.
2671 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2672 fixed by 4.65/MH/01 (is this a record?) are fixed:
2674 (i) An empty string was always treated as zero by the numeric comparison
2675 operators. This behaviour has been restored.
2677 (ii) It is documented that the numeric comparison operators always treat
2678 their arguments as decimal numbers. This was broken in that numbers
2679 starting with 0 were being interpreted as octal.
2681 While fixing these problems I realized that there was another issue that
2682 hadn't been noticed. Values of message_size_limit (both the global option
2683 and the transport option) were treated as octal if they started with 0.
2684 The documentation was vague. These values are now always treated as
2685 decimal, and I will make that clear in the documentation.
2691 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2692 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2695 MH/01 Don't check that the operands of numeric comparison operators are
2696 integers when their expansion is in "skipping" mode (fixes bug
2697 introduced by 4.64-PH/07).
2699 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2700 child addresses, Exim now panics and dies. Previously, because the count
2701 is held in a short int, deliveries were likely to be lost. As such a
2702 large number of recipients for a single message is ridiculous
2703 (performance will be very, very poor), I have chosen to impose a limit
2704 rather than extend the field.
2710 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2711 leftover -K file (the existence of which was triggered by #402).
2712 While we were at it, introduced process PID as part of the -K
2713 filename. This should rule out race conditions when creating
2716 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2717 processing considerably. Previous code took too long for large mails,
2718 triggering a timeout which in turn triggers #401.
2720 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2721 in the DK code in transports.c. sendfile() is not really portable,
2722 hence the _LINUX specificness.
2724 TF/01 In the add_headers option to the mail command in an Exim filter,
2725 there was a bug that Exim would claim a syntax error in any
2726 header after the first one which had an odd number of characters
2729 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2730 callout verification, Exim cached a "reject" for the entire domain. This
2731 is correct for most verifications, but it is not correct for a recipient
2732 verification with use_sender or use_postmaster set, because in that case
2733 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2734 case of MAIL FROM:<> rejection from other early rejections (e.g.
2735 rejection of HELO). When verifying a recipient using a non-null MAIL
2736 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2737 Whatever the result of the callout, the value of the domain cache is
2738 left unchanged (for any other kind of callout, getting as far as trying
2739 RCPT means that the domain itself is ok).
2741 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2744 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2745 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2746 session, the daemon ignores ECONNECT errors and logs others; it now
2747 ignores EPIPE as well.
2749 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2750 (quoted-printable decoding).
2752 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2753 later the small subsequent patch to fix an introduced bug.
2755 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2757 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2759 PH/08 An error is now given if message_size_limit is specified negative.
2761 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2762 to be given (somewhat) arbitrary names.
2764 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2767 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2768 miscellaneous code fixes
2770 PH/10 Added the log_reject_target ACL modifier to specify where to log
2773 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2774 hostname. This is wrong, because it relates to the incoming message (and
2775 probably the interface on which it is arriving) and not to the outgoing
2776 callout (which could be using a different interface). This has been
2777 changed to use the value of the helo_data option from the smtp transport
2778 instead - this is what is used when a message is actually being sent. If
2779 there is no remote transport (possible with a router that sets up host
2780 addresses), $smtp_active_hostname is used.
2782 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2783 tweaks were necessary in order to get it to work (see also 21 below):
2784 (a) The code assumed that strncpy() returns a negative number on buffer
2785 overflow, which isn't the case. Replaced with Exim's string_format()
2787 (b) There were several signed/unsigned issues. I just did the minimum
2788 hacking in of casts. There is scope for a larger refactoring.
2789 (c) The code used strcasecmp() which is not a standard C function.
2790 Replaced with Exim's strcmpic() function.
2791 (d) The code set only $1; it now sets $auth1 as well.
2792 (e) A simple test gave the error "authentication client didn't specify
2793 service in request". It would seem that Dovecot has changed its
2794 interface. Fortunately there's a specification; I followed it and
2795 changed what the client sends and it appears to be working now.
2797 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2800 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2801 address A is aliased to B and C, where B exists and C does not. Without
2802 -v the output is "A verified" because verification stops after a
2803 successful redirection if more than one address is generated. However,
2804 with -v the child addresses are also verified. Exim was outputting "A
2805 failed to verify" and then showing the successful verification for C,
2806 with its parentage. It now outputs "B failed to verify", showing B's
2807 parentage before showing the successful verification of C.
2809 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2810 look up a TXT record in a specific list after matching in a combined
2813 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2814 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2815 they consult the DNS. I had assumed they would set it the way they
2816 wanted; and indeed my experiments on Linux seem to show that in some
2817 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2818 To be on the safe side, however, I have now made the interface to
2819 host_find_byname() similar to host_find_bydns(), with an argument
2820 containing the DNS resolver options. The host_find_byname() function now
2821 sets these options at its start, just as host_find_bydns() does. The smtp
2822 transport options dns_qualify_single and dns_search_parents are passed to
2823 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2824 of host_find_byname() use the default settings of RES_DEFNAMES
2825 (qualify_single) but not RES_DNSRCH (search_parents).
2827 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2828 spool_read_header() do less string testing, by means of a preliminary
2829 switch on the second character of optional "-foo" lines. (This is
2830 overdue, caused by the large number of possibilities that now exist.
2831 Originally there were few.) While I was there, I also converted the
2832 str(n)cmp tests so they don't re-test the leading "-" and the first
2833 character, in the hope this might squeeze out yet more improvement.
2835 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2836 flag allowing group syntax was set by the header_syntax check but not
2837 turned off, possible causing trouble later; (2) The flag was not being
2838 set at all for the header_verify test, causing "group"-style headers to
2839 be rejected. I have now set it in this case, and also caused header_
2840 verify to ignore an empty address taken from a group. While doing this, I
2841 came across some other cases where the code for allowing group syntax
2842 while scanning a header line wasn't quite right (mostly, not resetting
2843 the flag correctly in the right place). These bugs could have caused
2844 trouble for malformed header lines. I hope it is now all correct.
2846 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2847 with the "reply" argument non-NULL. The code, however (which originally
2848 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2849 but it didn't always do it. This confused somebody who was copying the
2850 code for some other use. I have removed all the tests.
2852 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2853 feature that was used to support insecure browsers during the U.S. crypto
2854 embargo. It requires special client support, and Exim is probably the
2855 only MTA that supported it -- and would never use it because real RSA is
2856 always available. This code has been removed, because it had the bad
2857 effect of slowing Exim down by computing (never used) parameters for the
2858 RSA_EXPORT functionality.
2860 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2861 authenticator to fail if there's a tab character in the incoming data
2862 (there should never be unless someone is messing about, as it's supposed
2863 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2864 if the connection is using TLS or if the remote IP is the same as the
2865 local IP, and the "valid-client-cert option" if a client certificate has
2868 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2869 authenticators. This can be used for authorization after authentication
2870 succeeds. (In the case of plaintext, it servers for both authentication
2873 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2874 if any retry times were supplied.
2876 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2877 connection, where there is no client IP address to check. In this
2878 situation, the verify now always succeeds.
2880 PH/25 Applied John Jetmore's -Mset patch.
2882 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2884 PH/27 In a string expansion for a processed (not raw) header when multiple
2885 headers of the same name were present, leading whitespace was being
2886 removed from all of them, but trailing whitespace was being removed only
2887 from the last one. Now trailing whitespace is removed from each header
2888 before concatenation. Completely empty headers in a concatenation (as
2889 before) are ignored.
2891 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2892 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2894 PH/29 [Removed. This was a change that I later backed out, and forgot to
2895 correct the ChangeLog entry (that I had efficiently created) before
2896 committing the later change.]
2898 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2899 address errors (4xx response to RCPT) over the same connection as other
2900 messages routed to the same hosts. Such deliveries are always "forced",
2901 so retry times are not inspected. This resulted in far too many retries
2902 for the affected addresses. The effect occurred only when there were more
2903 hosts than the hosts_max_try setting in the smtp transport when it had
2904 the 4xx errors. Those hosts that it had tried were not added to the list
2905 of hosts for which the message was waiting, so if all were tried, there
2906 was no problem. Two fixes have been applied:
2908 (i) If there are any address or message errors in an SMTP delivery, none
2909 of the hosts (tried or untried) are now added to the list of hosts
2910 for which the message is waiting, so the message should not be a
2911 candidate for sending over the same connection that was used for a
2912 successful delivery of some other message. This seems entirely
2913 reasonable: after all the message is NOT "waiting for some host".
2914 This is so "obvious" that I'm not sure why it wasn't done
2915 previously. Hope I haven't missed anything, but it can't do any
2916 harm, as the worst effect is to miss an optimization.
2918 (ii) If, despite (i), such a delivery is accidentally attempted, the
2919 routing retry time is respected, so at least it doesn't keep
2920 hammering the server.
2922 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2923 in ${readsocket because some servers need this prod.
2925 PH/32 Added some extra debug output when updating a wait-xxx database.
2927 PH/33 The hint "could be header name not terminated by colon", which has been
2928 given for certain expansion errors for a long time, was not being given
2929 for the ${if def:h_colon_omitted{... case.
2931 PH/34 The spec says: "With one important exception, whenever a domain list is
2932 being scanned, $domain contains the subject domain." There was at least
2933 one case where this was not true.
2935 PH/35 The error "getsockname() failed: connection reset by peer" was being
2936 written to the panic log as well as the main log, but it isn't really
2937 panic-worthy as it just means the connection died rather early on. I have
2938 removed the panic log writing for the ECONNRESET error when getsockname()
2941 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2942 runs only) independently of the message's sender address. This meant
2943 that, if the 4xx error was in fact related to the sender, a different
2944 message to the same recipient with a different sender could confuse
2945 things. In particular, this can happen when sending to a greylisting
2946 server, but other circumstances could also provoke similar problems.
2947 I have changed the default so that the retry time for these errors is now
2948 based a combination of the sender and recipient addresses. This change
2949 can be overridden by setting address_retry_include_sender=false in the
2952 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2953 remote server are returned as part of bounce messages. This was not
2954 happening for LMTP over a pipe (the lmtp transport), but now it is the
2955 same for both kinds of LMTP.
2957 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2958 in header lines that were in fact CNAMEs. This is no longer the case.
2960 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2961 and queue runs started by the daemon processed all messages. This has
2962 been fixed so that -R and -S can now usefully be given with -q<time>.
2964 PH/40 Import PCRE release 6.7 (fixes some bugs).
2966 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2968 PH/42 Give an error if -q is specified more than once.
2970 PH/43 Renamed the variables $interface_address and $interface_port as
2971 $received_ip_address and $received_port, to make it clear that these
2972 values apply to message reception, and not to the outgoing interface when
2973 a message is delivered. (The old names remain recognized, of course.)
2975 PH/44 There was no timeout on the connect() call when using a Unix domain
2976 socket in the ${readsocket expansion. There now is.
2978 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2979 be meaningful with "accept".
2981 SC/01 Eximstats V1.43
2982 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2984 SC/02 Eximstats V1.44
2985 Use a glob alias rather than an array ref in the generated
2986 parser. This improves both readability and performance.
2988 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2989 Collect SpamAssassin and rejection statistics.
2990 Don't display local sender or destination tables unless
2991 there is data to show.
2992 Added average volumes into the top table text output.
2994 SC/04 Eximstats V1.46
2995 Collect data on the number of addresses (recipients)
2996 as well as the number of messages.
2998 SC/05 Eximstats V1.47
2999 Added 'Message too big' to the list of mail rejection
3000 reasons (thanks to Marco Gaiarin).
3002 SC/06 Eximstats V1.48
3003 Mainlog lines which have GMT offsets and are too short to
3004 have a flag are now skipped.
3006 SC/07 Eximstats V1.49 (Alain Williams)
3007 Added the -emptyok flag.
3009 SC/08 Eximstats V1.50
3010 Fixes for obtaining the IP address from reject messages.
3012 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3013 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3014 whitespace changes from 4.64-PH/27
3016 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3019 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3020 are found, allow negative numbers in numeric criteria)
3022 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3024 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3025 to match changes made in 4.64-PH/43
3027 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3029 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3030 "domain" is actually an IP address, and force a failure. This locks out
3031 those revolvers/nameservers that support "A-for-A" lookups, in
3032 contravention of the specifications.
3034 PH/48 When a host name was looked up from an IP address, and the subsequent
3035 forward lookup of the name timed out, the host name was left in
3036 $sender_host_name, contrary to the specification.
3038 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3039 restricted to single-key lookups, Exim was not diagnosing an error if
3040 * or *@ was used with a query-style lookup.
3042 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3044 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3045 long ago, but noone interested enough thought of it. Let's just say that
3046 the "1.1" means that there are some new functions that weren't there at
3047 some point in the past.
3049 PH/51 Error processing for expansion failure of helo_data from an smtp
3050 transport during callout processing was broken.
3052 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3053 tested/used via the -bh/-bhc/-bs options.
3055 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3056 bug, fixed in subsequent PCRE releases).
3058 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3059 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3061 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3067 SC/01 Use a glob alias rather than an array ref in eximstats generated
3068 parser. This improves both readability and performance.
3070 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3071 Don't display local sender or destination tables in eximstats unless
3072 there is data to show.
3073 Added average volumes into the eximstats top table text output.
3075 SC/03 Collect data on the number of addresses (recipients) as well
3076 as the number of messages in eximstats.
3078 TF/01 Correct an error in the documentation for the redirect router. Exim
3079 does (usually) call initgroups() when daemonizing.
3081 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3082 with consistent privilege compared to when running as a daemon.
3084 TF/03 Note in the spec that $authenticated_id is not set for local
3085 submissions from trusted users.
3087 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3088 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3090 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3091 by adding some example configuration directives to the default
3092 configuration file. A little bit of work is required to uncomment the
3093 directives and define how usernames and passwords are checked, but
3094 there is now a framework to start from.
3096 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3097 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3098 without this. I don't know how relevant this is to other LDAP libraries.
3100 PH/02 Add the verb name to the "unknown ACL verb" error.
3102 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3104 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3106 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3107 directory not expanded when it should be if an expanded home directory
3108 was set for the address (which is overridden by the transport).
3110 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3113 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3114 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3115 because it is too late at that time, and has no effect.
3117 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3118 security issue with \' (bugzilla #107). I could not use the
3119 PQescapeStringConn() function, because it needs a PGconn value as one of
3122 PH/08 When testing addresses using -bt, indicate those final addresses that
3123 are duplicates that would not cause an additional delivery. At least one
3124 person was confused, thinking that -bt output corresponded to deliveries.
3125 (Suppressing duplicates isn't a good idea as you lose the information
3126 about possibly different redirections that led to the duplicates.)
3128 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3129 systems where poll() doesn't work, in particular OS X.
3131 PH/10 Added more information to debugging output for retry time not reached.
3133 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3134 operations in malware.c.
3136 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3139 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3140 syslog_duplication set false, log lines that would normally be written
3141 both the the main log and to the reject log were not written to syslog at
3144 PH/14 In the default configuration, change the use of "message" in ACL warn
3145 statements to "add_header".
3147 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3148 not followed by a command (e.g. "seen endif").
3150 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3151 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3154 PH/17 Added extra conditions to the default value of delay_warning_condition
3158 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3159 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3160 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3163 The Auto-Submitted: and various List- headers are standardised, whereas I
3164 don't think Precedence: ever was.
3166 PH/18 Refactored debugging code in route_finduser() to show more information,
3167 in particular, the error code if getpwnam() issues one.
3169 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3170 This is apparently needed in addition to the PH/07 change above to avoid
3171 any possible encoding problems.
3173 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3174 but not after initializing Perl.
3176 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3177 output them only if debugging. By default they are written stderr,
3178 apparently, which is not desirable.
3180 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3183 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3186 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3188 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3189 authenticated or an ident call has been made. Suppress the default
3190 values for $authenticated_id and $authenticated_sender (but permit -oMai
3191 and -oMas) when testing with -bh.
3193 PH/24 Re-jigged the order of the tests in the default configuration so that the
3194 tests for valid domains and recipients precede the DNS black list and CSA
3195 tests, on the grounds that those ones are more expensive.
3197 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3198 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3199 This bug exists in every version of Exim that I still have, right back to
3202 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3203 However, an attempt to turn on case-sensitivity in a regex key by
3204 including (?-i) didn't work because the subject string was already
3205 lowercased, and the effects were non-intuitive. It turns out that a
3206 one-line patch can be used to allow (?-i) to work as expected.
3212 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3213 other effects) broke the use of negated acl sub-conditions.
3215 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3218 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3219 "Deny" causes Exim to reject the incoming connection with a 554 error.
3220 Unfortunately, if there is a major crisis, such as a disk failure,
3221 tcp-wrappers gives "deny", whereas what one would like would be some
3222 kind of temporary error. A kludge has been added to help with this.
3223 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3224 554 error is used if errno is still zero or contains ENOENT (which occurs
3225 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3228 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3230 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3231 errors. Otherwise a message that provokes a temporary error (when other
3232 messages do not) can cause a whole host to time out.
3234 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3235 addresses were routed directly to files or pipes from a redirect router.
3236 File deliveries just didn't batch; pipe deliveries might have suffered
3239 PH/06 A failure to get a lock for a hints database would erroneously always say
3240 "Failed to get write lock", even when it was really a read lock.
3242 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3243 of 0600. This has been changed to use the value of the lockfile_mode
3244 option (which defaults to 0600).
3246 PH/08 Applied small patch from the Sieve maintainer.
3248 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3249 folder from quota calculations, a direct delivery into this folder messed
3250 up the contents of the maildirsize file. This was because the regex was
3251 used only to exclude .Trash (or whatever) when the size of the mailbox
3252 was calculated. There was no check that a delivery was happening into an
3253 excluded directory. This bug has been fixed by ignoring all quota
3254 processing for deliveries into excluded directories.
3256 PH/10 Added the maildirfolder_create_regex option to appendfile.
3262 PH/01 The code for finding all the local interface addresses on a FreeBSD
3263 system running IPv6 was broken. This may well have applied to all BSD
3264 systems, as well as to others that have similar system calls. The broken
3265 code found IPv4 interfaces correctly, but gave incorrect values for the
3266 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3267 that it would not match correctly against @[] and not recognize the IPv6
3270 PH/02 The ipliteral router was not recognizing addresses of the form user@
3271 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3273 PH/03 Added disable_ipv6.
3275 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3276 decoded form, because it is most often used to construct To: headers
3277 lines in autoreplies, and the decoded form may well be syntactically
3278 invalid. However, $reply_address has leading white space removed, and all
3279 newlines turned into spaces so that the autoreply transport does not
3282 PH/05 If group was specified without a user on a router, and no group or user
3283 was specified on a transport, the group from the router was ignored.
3285 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3286 for visible compile-time settings that can be used to change these
3287 numbers, for those that want even more. Backwards compatibility with old
3288 spool files has been maintained. However, going back to a previous Exim
3289 release will lost any variables that are in spool files.
3291 PH/07 Two small changes when running in the test harness: increase delay when
3292 passing a TCP/IP connection to a new process, in case the original
3293 process has to generate a bounce, and remove special handling of
3294 127.0.0.2 (sic), which is no longer necessary.
3296 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3297 be the same on different OS.
3299 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3302 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3303 whether --show-vars was specified or not
3305 JJ/02 exipick: Added support for new ACL variable spool format introduced
3308 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3309 syntactically invalid From: or Reply-to: line, and a filter used this to
3310 generate an autoreply, and therefore failed to obtain an address for the
3311 autoreply, Exim could try to deliver to a non-existent relative file
3312 name, causing unrelated and misleading errors. What now happens is that
3313 it logs this as a hard delivery error, but does not attempt to create a
3316 PH/11 The exinext utility has a -C option for testing purposes, but although
3317 the given file was scanned by exinext itself; it wasn't being passed on
3318 when Exim was called.
3320 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3321 an end-of-file indication when reading a command response.
3323 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3324 compiled. In many other places in Exim, IPv6 addresses are always
3325 recognized, so I have changed this. It also means that IPv4 domain
3326 literals of the form [IPV4:n.n.n.n] are now always recognized.
3328 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3329 used if the router is not running as root, for example, when verifying at
3330 ACL time, or when using -bh. The debugging output from this situation was
3331 non-existent - all you got was a failure to exec. I have made two
3334 (a) Failures to set uid/gid, the current directory, or a process leader
3335 in a subprocess such as that created by queryprogram now generate
3336 suitable debugging output when -d is set.
3338 (b) The queryprogram router detects when it is not running as root,
3339 outputs suitable debugging information if -d is set, and then runs
3340 the subprocess without attempting to change uid/gid.
3342 PH/15 Minor change to Makefile for building test_host (undocumented testing
3345 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3346 additional section of a DNS packet that returns MX or SRV records.
3347 Instead, it always explicitly searches for A/AAAA records. This avoids
3348 major problems that occur when a DNS server includes only records of one
3349 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3350 fixed another bug: if SRV records were looked up and the corresponding
3351 address records were *not* found in the additional section, the port
3352 values from the SRV records were lost.
3354 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3355 using the correct key (the original address) when searching the retry
3356 rules in order to find which one to use for generating the retry hint.
3358 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3359 adding the default one. Similarly, if it contains a Reply-To: header, the
3360 errors_reply_to option, if set, is not used.
3362 PH/19 When calculating a retry time, Exim used to measure the "time since
3363 failure" by looking at the "first failed" field in the retry record. Now
3364 it does not use this if it is later than than the arrival time of the
3365 message. Instead it uses the arrival time. This makes for better
3366 behaviour in cases where some deliveries succeed, thus re-setting the
3367 "first failed" field. An example is a quota failure for a huge message
3368 when small messages continue to be delivered. Without this change, the
3369 "time since failure" will always be short, possible causing more frequent
3370 delivery attempts for the huge message than are intended.
3371 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3373 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3374 $1, $2, $3) because the numerical variables can be reset during some
3375 expansion items (e.g. "match"), thereby losing the authentication data.
3377 PH/21 Make -bV show the size of off_t variables so that the test suite can
3378 decide whether to run tests for quotas > 2G.
3380 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3381 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3382 filecount value is greater than 2G or if a quota value is greater than 2G
3383 on a system where the size of off_t is not greater than 4, a panic error
3386 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3387 never match. The debug and -bh output now contains an explicit error
3388 message indicating a malformed IPv4 address or mask.
3390 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3391 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3392 PH/23 above applies.
3394 PH/25 Do not write to syslog when running in the test harness. The only
3395 occasion when this arises is a failure to open the main or panic logs
3396 (for which there is an explicit test).
3398 PH/26 Added the /no_tell option to "control=freeze".
3400 PH/27 If a host name lookup failed very early in a connection, for example, if
3401 the IP address matched host_lookup and the reverse lookup yielded a name
3402 that did not have a forward lookup, an error message of the form "no IP
3403 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3404 could be logged. Now it outputs the IP address instead of "NULL".
3406 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3407 allows the sender and the authenticated sender to be set when
3408 submitting a message from within Exim. Since child_open_exim() is
3409 documented for local_scan(), the new function should be too.
3411 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3412 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3413 results in an empty string is now treated as unset.
3415 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3417 PH/31 Added sender_verify_fail logging option.
3419 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3420 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3421 tidied the source and removed it altogether.
3423 PH/33 When a queue run was abandoned because the load average was too high, a
3424 log line was always written; now it is written only if the queue_run log
3425 selector is set. In addition, the log line for abandonment now contains
3426 information about the queue run such as the pid. This is always present
3427 in "start" and "stop" lines but was omitted from the "abandon" line.
3429 PH/34 Omit spaces between a header name and the colon in the error message that
3430 is given when verify = headers_syntax fails (if there are lots of them,
3431 the message gets confusing).
3433 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3434 names, as there are now some PTR records that contain slashes. This check
3435 is only to protect against broken name servers that fall over on strange
3436 characters, so the fact that it applies to all lookups doesn't matter.
3438 PH/36 Now that the new test suite is complete, we can remove some of the
3439 special code in Exim that was needed for the old test suite. For example,
3440 sorting DNS records because real resolvers return them in an arbitrary
3441 order. The new test suite's fake resolver always returns records in the
3444 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3445 bounces) except when queue_only is set, to avoid logging races between
3446 the different processes.
3448 PH/38 Panic-die if .include specifies a non-absolute path.
3450 PH/39 A tweak to the "H" retry rule from its user.
3452 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3453 a label. They prevented compilation on older perls.
3455 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3456 a warning to be raised on newish perls.
3458 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3459 on queue. Changes to match documented behaviour of showing count of
3460 messages matching specified criteria.
3462 PH/40 Changed the default ident timeout from 30s to 5s.
3464 PH/41 Added support for the use of login_cap features, on those BSD systems
3465 that have them, for controlling the resources used by pipe deliveries.
3467 PH/42 The content-scanning code uses fopen() to create files in which to put
3468 message data. Previously it was not paying any attention to the mode of
3469 the files. Exim runs with umask(0) because the rest of the code creates
3470 files with open(), and sets the required mode explicitly. Thus, these
3471 files were ending up world-writeable. This was not a big issue, because,
3472 being within the spool directory, they were not world-accessible. I have
3473 created a function called modefopen, which takes an additional mode
3474 argument. It sets umask(777), creates the file, chmods it to the required
3475 mode, then resets the umask. All the relevant calls to fopen() in the
3476 content scanning code have been changed to use this function.
3478 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3479 to 24 hours. This avoids potential overflow problems when processing G
3480 and H retry rules. I suspect nobody ever tinkers with this value.
3482 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3484 PH/45 When the plaintext authenticator is running as a client, the server's
3485 challenges are checked to ensure they are valid base64 strings. By
3486 default, the authentication attempt is cancelled if an invalid string is
3487 received. Setting client_ignore_invalid_base64 true ignores these errors.
3488 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3489 they are received. Thus, the responses can be made to depend on the
3490 challenges. If an invalid string is ignored, an empty string is placed in
3493 PH/46 Messages that are created by the autoreply transport now contains a
3494 References: header, in accordance with RFCs 2822 and 3834.
3496 PH/47 Added authenticated_sender_force to the smtp transport.
3498 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3500 PH/49 Installed latest patch from the Sieve maintainer.
3502 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3503 was also set, the appendfile transport was unnecessarily scanning a
3504 directory of message files (e.g. for maildir delivery) to find the count
3505 of files (along with the size), even though it did not need this
3506 information. It now does the scan only if it needs to find either the
3507 size of the count of files.
3509 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3511 PH/52 Two bugs concerned with error handling when the smtp transport is
3514 (i) Exim was not creating retry information for temporary errors given
3515 for individual recipients after the DATA command when the smtp transport
3516 was used in LMTP mode. This meant that they could be retried too
3517 frequently, and not timed out correctly.
3519 (ii) Exim was setting the flag that allows error details to be returned
3520 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3521 recipients that were returned after the DATA command.
3523 PH/53 This is related to PH/52, but is more general: for any failing address,
3524 when detailed error information was permitted to be returned to the
3525 sender, but the error was temporary, then after the final timeout, only
3526 "retry timeout exceeded" was returned. Now it returns the full error as
3527 well as "retry timeout exceeded".
3529 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3530 do this, and (what is worse) MTAs that accept it.
3532 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3533 will now be deprecated.
3535 PH/56 New os.c-cygwin from the Cygwin maintainer.
3537 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3538 formats (previously only available in exim formats via -bpr, -bpru,
3539 and -bpra. Now also available in native and exiqgrep formats)
3541 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3542 with very large, slow to parse queues
3544 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3546 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3548 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3549 responses to authentication challenges, though it was showing the
3550 challenges; (ii) I've removed the CR characters from the debug output for
3553 PH/58 Allow for the insertion of a newline as well as a space when a string
3554 is turned into more than one encoded-word during RFC 2047 encoding. The
3555 Sieve code now uses this.
3557 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3558 data_4xx, lost_connection, tls_required.
3560 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3561 message was being sent as an SMTP response.
3563 PH/61 Add -l and -k options to exicyclog.
3565 PH/62 When verifying, if an address was redirected to one new address, so that
3566 verification continued, and the new address failed or deferred after
3567 having set something in $address_data, the value of $address_data was not
3568 passed back to the ACL. This was different to the case when no
3569 redirection occurred. The value is now passed back in both cases.
3571 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3572 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3573 use login_cap.h, so on its own it isn't the distinguishing feature. The
3574 new name refers directly to the setclassresources() function.
3576 PH/65 Added configuration files for NetBSD3.
3578 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3580 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3581 is preferred over IPv4.
3583 PH/68 The bounce_return_message and bounce_return_body options were not being
3584 honoured for bounces generated during the reception of non-SMTP messages.
3585 In particular, this applied to messages rejected by the ACL. This bug has
3586 been fixed. However, if bounce_return_message is true and bounce_return_
3587 body is false, the headers that are returned for a non-SMTP message
3588 include only those that have been read before the error was detected.
3589 (In the case of an ACL rejection, they have all been read.)
3591 PH/69 The HTML version of the specification is now built in a directory called
3592 spec_html instead of spec.html, because the latter looks like a path with
3593 a MIME-type, and this confuses some software.
3595 PH/70 Catch two compiler warnings in sieve.c.
3597 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3598 function verify_get_ident() calls ip_connect() to connect a socket, but
3599 if the "connect()" function timed out, ip_connect() used to close the
3600 socket. However, verify_get_ident() also closes the socket later, and in
3601 between Exim writes to the log, which may get opened at this point. When
3602 the socket was closed in ip_connect(), the log could get the same file
3603 descriptor number as the socket. This naturally causes chaos. The fix is
3604 not to close the socket in ip_connect(); the socket should be closed by
3605 the function that creates it. There was only one place in the code where
3606 this was missing, in the iplookup router, which I don't think anybody now
3607 uses, but I've fixed it anyway.
3609 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3610 well as to direct DNS lookups. Otherwise the handling of names in host
3611 lists is inconsistent and therefore confusing.
3617 PH/01 Two changes to the default runtime configuration:
3619 (1) Move the checks for relay_from_hosts and authenticated clients from
3620 after to before the (commented out) DNS black list checks.
3622 (2) Add control=submission to the relay_from_hosts and authenticated
3623 clients checks, on the grounds that messages accepted by these
3624 statements are most likely to be submissions.
3626 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3628 (1) Generate an error if the third argument for the ${prvs expansion is
3631 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3634 (3) Reset the variables that are obtained from the first argument of
3635 ${prvscheck and used in the second argument before leaving the code,
3636 because their memory is reclaimed, so using them afterwards may do
3639 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3640 one (it's much easier than Tom thought :-).
3642 (5) Because of (4), we can now allow for the use of $prvscheck_result
3643 inside the third argument.
3645 PH/03 For some reason, the default setting of PATH when running a command from
3646 a pipe transport was just "/usr/bin". I have changed it to
3649 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3650 anything to be listed in the output from -bV.
3652 PH/05 When a filter generated an autoreply, the entire To: header line was
3653 quoted in the delivery log line, like this:
3655 => >A.N.Other <ano@some.domain> <original@ddress> ...
3657 This has been changed so that it extracts the operative address. There
3658 may be more than one such address. If so, they are comma-separated, like
3661 => >ano@some.domain,ona@other.domain <original@ddress> ...
3663 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3664 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3665 not being looked up in the rDNS to get a host name, Exim was showing the
3666 IP address twice in Received: lines, even though the IP addresses were
3667 identical. For example:
3669 Received: from [1.2.3.4] (helo=[1.2.3.4])
3671 However, if the real host name was known, it was omitting the HELO data
3672 if it matched the actual IP address. This has been tidied up so that it
3673 doesn't show the same IP address twice.
3675 PH/07 When both +timestamp and +memory debugging was on, the value given by
3676 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3677 called by the debug printing, thereby overwriting the timestamp buffer.
3678 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3680 PH/08 When the original message was included in an autoreply transport, it
3681 always said "this is a copy of the message, including all the headers",
3682 even if body_only or headers_only was set. It now gives an appropriate
3685 PH/09 Applied a patch from the Sieve maintainer which:
3687 o fixes some comments
3688 o adds the (disabled) notify extension core
3689 o adds some debug output for the result of if/elsif tests
3690 o points to the current vacation draft in the documentation
3691 and documents the missing references header update
3695 o fixes a bug in processing the envelope test (when testing
3696 multiple envelope elements, the last element determined the
3699 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3700 Electronic Mail") by including:
3702 Auto-submitted: auto-generated
3704 in the messages that it generates (bounce messages and others, such as
3705 warnings). In the case of bounce messages for non-SMTP messages, there was
3706 also a typo: it was using "Auto_submitted" (underscore instead of
3707 hyphen). Since every message generated by Exim is necessarily in response
3708 to another message, thes have all been changed to:
3710 Auto-Submitted: auto-replied
3712 in accordance with these statements in the RFC:
3714 The auto-replied keyword:
3716 - SHOULD be used on messages sent in direct response to another
3717 message by an automatic process,
3719 - MUST NOT be used on manually-generated messages,
3721 - MAY be used on Delivery Status Notifications (DSNs) and Message
3722 Disposition Notifications (MDNs),
3724 - MUST NOT be used on messages generated by automatic or periodic
3725 processes, except for messages which are automatic responses to
3728 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3729 to the default Received: header definition.
3731 PH/12 Added log selector acl_warn_skipped (default on).
3733 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3734 variables because (a) they are in the wrong storage pool and (b) even if
3735 they were copied, it wouldn't work properly because of the caching.
3737 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3738 checking when decoding. Apparently there are clients that generate
3739 overlong encoded strings. Why am I not surprised?
3741 PH/15 If the first argument of "${if match_address" was not empty, but did not
3742 contain an "@" character, Exim crashed. Now it writes a panic log message
3743 and treats the condition as false.
3745 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3747 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3748 extension "envelope-auth". The code is finished and in agreement with
3749 other implementations, but there is no documentation so far and in fact,
3750 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3751 not changing the active code.
3753 Print executed "if" and "elsif" statements when debugging is used. This
3754 helps a great deal to understand what a filter does.
3756 Document more things not specified clearly in RFC3028. I had all this
3757 sorted out, when out of a sudden new issues came to my mind. Oops."
3759 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3762 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3763 canonical form (as documented). However, after a host name lookup from
3764 the IP address, check_host() was doing a simple string comparison with
3765 addresses acquired from the DNS when checking that the found name did
3766 have the original IP as one of its addresses. Since any found IPv6
3767 addresses are likely to be in abbreviated form, the comparison could
3768 fail. Luckily, there already exists a function for doing the comparison
3769 by converting both addresses to binary, so now that is used instead of
3770 the text comparison.
3772 PH/20 There was another similar case to PH/19, when a complete host name was
3773 given in a host list; looking up its IP address could give an abbreviated
3774 form, whereas the current host's name might or might not be abbreviated.
3775 The same fix has been applied.
3781 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3782 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3785 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3786 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3788 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3790 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3791 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3792 Also, for "Auto-Submitted", treat anything other than "no" as
3793 non-personal, in accordance with RFC 3834. (Previously it treated
3794 anything starting "auto-" as non-personal.)
3796 TF/01 The control=submission/name=... option had a problem with syntax
3797 errors if the name included a slash character. The /name= option
3798 now slurps the rest of the string, so it can include any characters
3799 but it must come last in the list of options (after /sender_retain
3802 PH/05 Some modifications to the interface to the fake nameserver for the new
3810 TK/01 Added the "success_on_redirect" address verification option. See
3811 NewStuff for rationale and an example.
3813 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3815 PH/02 Patch to exigrep to allow it to work on syslog lines.
3817 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3818 fread() to skip over the body file's header line, because in Cygwin the
3819 header line is locked and is inaccessible.
3821 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3822 co-exist for some time) to make it clear that it is the Exim ID that is
3823 referenced, not the Message-ID: header line.
3825 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3826 string_format() function, because snprintf() does not exist on all
3829 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3830 ${stat: expansion item.
3832 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3833 protocol synchronization error", to keep the pedants happy.
3835 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3836 well as for IRIX systems, when gcc is being used. See the host.c source
3839 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3841 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3844 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3845 transport and to the smtp transport in LMTP mode.
3847 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3849 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3850 run a filter in a subprocess. This could lead to confusion in subsequent
3851 lookups in the parent process. There should also be a search_tidyup() at
3852 the end of the subprocess.
3854 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3855 only if the host matched helo_try_verify_hosts, which caused the
3856 verification to occur when the EHLO/HELO command was issued. The ACL just
3857 tested the remembered result. Now, if a previous verification attempt has
3858 not happened, "verify = helo" does it there and then.
3860 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3862 TK/03 Fix log output including CR from clamd.
3864 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3865 exist provoked a memory error which could cause a segfault.
3867 PH/15 Installed PCRE 6.2
3869 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3871 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3872 of the problem. Specifically, suggested +O2 rather than +O1 for the
3875 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3877 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3878 record, the port was not correctly logged when the outgoing_port log
3879 selector was set (it logged the transort's default port).
3881 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3882 fallback_hosts, and "hosts" in the smtp transport.
3884 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3885 host errors such as "Connection refused".
3887 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3888 authentication with radiusclient 0.4.9:
3890 - Error returned from rc_read_config was caught wrongly
3891 - Username/password not passed on to radius server due to wrong length.
3893 The presumption is that some radiusclient API changes for 4.51/PH/17
3894 were not taken care of correctly. The code is still untested by me (my
3895 Linux distribution still has 0.3.2 of radiusclient), but it was
3896 contributed by a Radius user.
3898 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3899 expanding the "port" option of the smtp transport.
3901 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3902 while reading a MIME header. Thanks to Tom Hughes for a patch.
3904 PH/24 Include config.h inside local_scan.h so that configuration settings are
3907 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3908 that in an ACL for RCPT (for example), you can examine exactly what was
3911 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3912 commands, but it was not correctly comparing the address with the actual
3913 client host address. Thus, it would show the EHLO address in Received:
3914 header lines when this was not necessary.
3916 PH/27 Added the % operator to ${eval:}.
3918 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3919 it should be ignoring failures (because with -C, for example, it has lost
3920 privilege). It wasn't ignoring creation failures other than "already
3923 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3926 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3927 because an input line was too long, either on its own, or by virtue of
3928 too many continuations, the temporary file was not being removed, and the
3929 return code was incorrect.
3931 PH/31 Missing "BOOL" in function definition in filtertest.c.
3933 PH/32 Applied Sieve patches from the maintainer.
3935 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3937 PH/33 Added "verify = not_blind".
3939 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3940 Local/Makefile (with some defaults set). These are used in built scripts
3941 such as exicyclog, but they have never been used in the exim_install
3942 script (though there are many overriding facilities there). I have
3943 arranged that the exim_install script now takes note of these two
3946 PH/35 Installed configuration files for Dragonfly.
3948 PH/36 When a locally submitted message by a trusted user did not contain a
3949 From: header, and the sender address was obtained from -f or from an SMTP
3950 MAIL command, and the trusted user did not use -F to supply a sender
3951 name, $originator_name was incorrectly used when constructing a From:
3952 header. Furthermore, $originator_name was used for submission mode
3953 messages from external hosts without From: headers in a similar way,
3954 which is clearly wrong.
3956 PH/37 Added control=suppress_local_fixups.
3958 PH/38 When log_selector = +received_sender was set, and the addition of the
3959 sender made the log line's construction buffer exactly full, or one byte
3960 less than full, an overflow happened when the terminating "\n" was
3963 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3964 when the result of a list match is failure because a DNS lookup failed.
3966 PH/40 RM_COMMAND is now used in the building process.
3968 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3969 the "build-* directories that it finds.
3971 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3972 address in a domain literal was a prefix of an interface address.
3974 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3975 when verifying a sender address, unless rewrite_headers is false.
3977 PH/44 Wrote a long comment about why errors_to addresses are verified as
3978 recipients, not senders.
3980 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3981 the ratelimit ACL was added.
3983 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3985 PH/46 Added extra information about PostgreSQL errors to the error string.
3987 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3988 avoiding the need to install special zones in a real server. This is
3989 backwards compatible; if it can't find the fake resolver, it drops back.
3990 Thus, both old and new test suites can be run.
3992 TF/02 Added util/ratelimit.pl
3994 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3995 clock is set back in time.
3997 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3998 Candler <B.Candler@pobox.com>.
4000 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4001 OK for addresses that are the result of redirections.
4003 PH/48 A number of further additions for the benefit of the new test suite,
4004 including a fake gethostbyname() that interfaces to the fake DNS resolver
4007 TF/06 The fix for widen_domains has also been applied to qualify_single and
4008 search_parents which are the other dnslookup options that can cause
4011 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4014 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4016 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4017 service" if it encountered an MX record with an empty target hostname.
4018 The message is now "an MX or SRV record indicated no SMTP service".
4020 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4021 if the DNS of the sending site is misconfigured. This is quite a
4022 common situation. This change restores the behaviour of treating a
4023 helo verification defer as a failure.
4025 PH/51 If self=fail was set on a router, the bounce message did not include the
4026 actual error message.
4032 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4034 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4035 command itself ended in error, the underlying message about the transport
4036 filter timeout was being overwritten with the pipe command error. Now the
4037 underlying error message should be appended to the second error message.
4039 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4041 PH/02 Reduce the amount of output that "make" produces by default. Full output
4042 can still be requested.
4044 PH/03 The warning log line about a condition test deferring for a "warn" verb
4045 was being output only once per connection, rather than after each
4046 occurrence (because it was using the same function as for successful
4047 "warn" verbs). This seems wrong, so I have changed it.
4049 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4050 it should not have, which might have caused a crash in the right
4051 circumstances, but probably never did.
4053 PH/04 Installed a modified version of Tony Finch's patch to make submission
4054 mode fix the return path as well as the Sender: header line, and to
4055 add a /name= option so that you can make the user's friendly name appear
4058 TF/03 Added the control = fakedefer ACL modifier.
4060 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4061 Mark Lowes for thorough testing.
4063 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4065 TK/03 Merged latest SRS patch from Miles Wilton.
4067 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4068 with the definition in sysexits.h (which is #included earlier).
4069 Fortunately, Exim does not actually use EX_OK. The code used to try to
4070 preserve the sysexits.h value, by assuming that macro definitions were
4071 scanned for macro replacements. I have been disabused of this notion,
4072 so now the code just undefines EX_OK before #including unistd.h.
4074 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4075 in the smtp transport. When a block could not be written in a single
4076 write() function, the timeout was being re-applied to each part-write.
4077 This seems wrong - if the receiver was accepting one byte at a time it
4078 would take for ever. The timeout is now adjusted when this happens. It
4079 doesn't have to be particularly precise.
4081 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4082 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4084 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4085 without a domain if the check to <postmaster@domain> fails.
4087 SC/01 Eximstats: added -xls and the ability to specify output files
4088 (patch written by Frank Heydlauf).
4090 SC/02 Eximstats: use FileHandles for outputting results.
4092 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4094 SC/04 Eximstats: fixed display of large numbers with -nvr option
4096 SC/05 Eximstats: fixed merging of reports with empty tables.
4098 SC/06 Eximstats: added the -include_original_destination flag
4100 SC/07 Eximstats: removed tabs and trailing whitespace.
4102 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4104 TK/06 MBOX spool code: Add real "From " MBOX separator line
4105 so the .eml file is really in mbox format (even though
4106 most programs do not really care). Patch from Alex Miller.
4108 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4109 The latter is generated from $received_to and is only set if the
4110 message has one envelope recipient. SA can use these headers,
4111 obviously out-of-the-box. Patch from Alex Miller.
4113 PH/08 The ${def test on a variable was returning false if the variable's
4114 value was "0", contrary to what the specification has always said!
4115 The result should be true unless the variable is empty.
4117 PH/09 The syntax error of a character other than { following "${if
4118 def:variable_name" (after optional whitespace) was not being diagnosed.
4119 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4120 accidental colon was present, for example, could give incorrect results.
4122 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4123 result is used (not including appendfile, where other changes are about
4126 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4127 This involved changing a lot of size variables from int to off_t. It
4128 should work with maildirs and everything.
4130 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4131 spamd dying while we are connected to it.
4133 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4136 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4137 function for BDB 4.3.
4139 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4141 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4142 the sources for PCRE have been re-arranged and more files are now
4145 PH/15 The code I had for printing potentially long long variables in PH/11
4146 above was not the best (it lost precision). The length of off_t variables
4147 is now inspected at build time, and an appropriate printing format (%ld
4148 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4149 to be "long long int" or "long int". This is needed for the internal
4150 formatting function string_vformat().
4152 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4153 the configuration file to be ":syslog", then the script "guesses" where
4154 the logs files are, rather than using the compiled in default. In our
4155 case the guess is not the same as the compiled default, so the script
4156 suddenly stopped working when I started to use syslog. The patch checks
4157 to see if log_file_path is "". If so, it attempts to read it from exim
4158 with no configuration file to get the compiled in version, before it
4159 falls back to the previous guessing code."
4161 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4162 implementing BATV in an Exim configuration. See NewStuff for the gory
4165 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4166 Makefile that are specific to HP-UX.
4168 PH/18 If the "use_postmaster" option was set for a recipient callout together
4169 with the "random" option, the postmaster address was used as the MAIL
4170 FROM address for the random test, but not for the subsequent recipient
4171 test. It is now used for both.
4173 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4174 patch removes a few documentation additions to RFC 3028, because the
4175 latest draft now contains them. It adds the new en;ascii-case comparator
4176 and a new error check for 8bit text in MIME parts. Comparator and
4177 require names are now matched exactly. I enabled the subaddress
4178 extension, but it is not well tested yet (read: it works for me)."
4180 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4181 rework some of the code of TK/09 above to avoid the hardwired use of
4182 "%lld" and "long long". Replaced the call to snprintf() with a call to
4185 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4186 records point to non-existent hosts", "retry timeout exceeded", and
4187 "retry time not reached for any host after a long failure period".
4189 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4190 experimental DomainKeys support:
4192 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4193 (2) On an error such as an illegally used "control", the wrong name for
4194 the control was given.
4196 These problems did NOT occur unless DomainKeys support was compiled.
4198 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4200 PH/24 Added ${if match_ip condition.
4202 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4203 databases so that it will be absolutely obvious if a crash occurs in the
4204 DB library. This is a regular occurrence (often caused by mis-matched
4207 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4208 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4209 picked up on a user's system that detects such things. There doesn't seem
4210 to be a gcc warning option for this - only an attribute that has to be
4211 put on the function's prototype. It seems that in Fedora Core 4 they have
4212 set this on a number of new functions. No doubt there will be more in due
4215 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4216 specify a transport. However, if an address that was verified by such a
4217 router was the subject of a callout, Exim crashed because it tried to
4218 read the rcpt_include_affixes from the non-existent transport. Now it
4219 just assumes that the setting of that option is false. This bug was
4220 introduced by 4.51/PH/31.
4222 PH/28 Changed -d+all to exclude +memory, because that information is very
4223 rarely of interest, but it makes the output a lot bigger. People tend to
4224 do -d+all out of habit.
4226 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4227 code in os-type was giving problems when libc.so lives in lib64, like on
4230 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4231 aren't the modern standard, and it seems that some systems' include files
4232 don't always have them. Exim was already checking for some of the newer
4233 ones like T_AAAA, and defining it itself. I've added checks for all the
4234 record types that Exim uses.
4236 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4237 not automatically generating a new one, as it is supposed to. This
4238 prevented TLS from working. If the file did exist, but contained invalid
4239 data, a new version was generated, as expected. It was only the case of a
4240 non-existent file that was broken.
4242 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4243 with a change in libdomainkeys > 0.64.
4245 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4246 from DNS. If the selector record carries the flag, it now has
4247 precedence over the domain-wide flag.
4249 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4251 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4252 the use of a port name that isn't defined in /etc/services) occurred, the
4253 message was deferred as in a normal delivery, and thus remained on the
4254 spool, instead of being failed because of the mua_wrapper setting. This
4255 is now fixed, and I tidied up some of the mua_wrapper messages at the
4258 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4259 the messages in a hash of arrays rather than using individual hashes.
4260 This is a bit cleaner and results in dramatic memory savings, albeit
4261 at a slight CPU cost.
4263 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4264 as requested by Marc Sherman.
4266 SC/10 Eximstats: added histograms for user specified patterns as requested
4269 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4271 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4272 fopen() in the content-scanning modules that did not already have it.
4278 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4279 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4281 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4283 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4285 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4286 to test Sieve filters that use "vacation".
4288 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4289 that changes the way the GnuTLS parameters are stored in the cache file.
4290 The new format can be generated externally. For backward compatibility,
4291 if the data in the cache doesn't make sense, Exim assumes it has read an
4292 old-format file, and it generates new data and writes a new file. This
4293 means that you can't go back to an older release without removing the
4296 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4297 work if there are any delivery delays because "one_time" forces the
4298 parent to be marked "delivered", so its unseen clone is never tried
4299 again. For this reason, Exim now forbids the simultaneous setting of
4302 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4303 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4304 entry if you want to know the details. The fix, however, neglected to
4305 consider the case when local delivery batching is involved. The test for
4306 "previously delivered" was not happening when checking to see if an
4307 address could be batched with a previous (undelivered) one; under
4308 certain circumstances this could lead to multiple deliveries to the same
4311 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4312 in its include files, and this causes problems building Exim.
4314 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4315 header_syntax) but Exim was just ignoring anything given after a slash.
4316 In particular, this caused confusion with an attempt to use "verify =
4317 reverse_host_lookup/defer_ok". An error is now given when options are
4318 supplied for verify items that do not have them. (Maybe reverse_host_
4319 lookup should have a defer_ok option, but that's a different point.)
4321 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4322 defined by RFC 821) to 2048, because there were problems with some AUTH
4323 commands, and RFC 1869 says the size should be increased for extended
4324 SMTP commands that take arguments.
4326 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4329 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4330 "unknown" error; now it says that the functionality isn't in the binary.
4332 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4333 an address' error message when a string expansion fails (syntax or
4334 whatever). Otherwise the password may appear in the log. Following change
4335 PH/42 below, there is no longer a chance of it appearing in a bounce
4338 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4340 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4341 of its addresses was ever tried. (Bugzilla bug #2.)
4343 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4344 the result incorrectly in the debug output. (It correctly added a newline
4345 to what was transported.)
4347 TF/01 Added $received_time.
4349 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4350 commented out examples of how to interface to a virus scanner and to
4351 SpamAssassin. Also added commented examples of av_scanner and
4352 spamd_address settings.
4354 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4355 and controls are allowed in which ACLs. There were a couple of minor
4356 errors. Some of the entries in the conditions table (which is a table of
4357 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4358 negation of where the condition IS allowed.
4360 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4362 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4363 header file does not have a version number, so I've had to invent a new
4364 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4365 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4366 radiusclient), but it was contributed by a Radius user.
4368 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4369 files or directories, for OpenSSL.
4371 PH/19 When an Exim process that is running as root has to create an Exim log
4372 file, it does so in a subprocess that runs as exim:exim so as to get the
4373 ownership right at creation (otherwise, other Exim processes might see
4374 the file with the wrong ownership). There was no test for failure of this
4375 fork() call, which would lead to the process getting stuck as it waited
4376 for a non-existent subprocess. Forks do occasionally fail when resources
4377 run out. I reviewed all the other calls to fork(); they all seem to check
4380 PH/20 When checking for unexpected SMTP input at connect time (before writing
4381 the banner), Exim was not dealing correctly with a non-positive return
4382 from the read() function. If the client had disconnected by this time,
4383 the result was a log entry for a synchronization error with an empty
4384 string after "input=" when read() returned zero. If read() returned -1
4385 (an event I could not check), uninitialized data bytes were printed.
4386 There were reports of junk text (parts of files, etc) appearing after
4389 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4391 PH/22 Added support for macro redefinition, and (re)definition in between
4392 driver and ACL definitions.
4394 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4395 forgetting to use the resulting value; it was using the unexpanded value.
4397 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4398 hadn't been configured. The fix is from Juergen Kreileder, who
4399 understands it better than I do:
4401 "Here's what I see happening with three configured cyrus_sasl
4402 authenticators configured (plain, login, cram-md5):
4404 On startup auth_cyrus_sasl_init() gets called for each of these.
4405 This means three calls to sasl_listmech() without a specified mech_list.
4406 => SASL tests which mechs of all available mechs actually work
4407 => three warnings about OTP not working
4408 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4410 With the patch, sasl_listmech() also gets called three times. But now
4411 SASL's mech_list option is set to the server_mech specified in the the
4412 authenticator. Or in other words, the answer from sasl_listmech()
4413 gets limited to just the mech you're testing for (which is different
4415 => the return list contains just 'plain' or 'login', 'cram-md5' or
4416 nothing depending on the value of ob->server_mech.
4418 I've just tested the patch: Authentication still works fine,
4419 unavailable mechs specified in the exim configuration are still
4420 caught, and the auth.log warnings about OTP are gone."
4422 PH/25 When debugging is enabled, the contents of the command line are added
4423 to the debugging output, even when log_selector=+arguments is not
4426 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4427 answer is "GNU", and only if the return is "GNU/something" is the answer
4430 PH/27 $acl_verify_message is now set immediately after the failure of a
4431 verification in an ACL, and so is available in subsequent modifiers. In
4432 particular, the message can be preserved by coding like this:
4434 warn !verify = sender
4435 set acl_m0 = $acl_verify_message
4437 Previously, $acl_verify_message was set only while expanding "message"
4438 and "log_message" when a very denied access.
4440 PH/28 Modified OS/os.c-Linux with
4442 -#ifndef OS_LOAD_AVERAGE
4443 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4445 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4446 nomenclature these days.)
4448 PH/29 Installed patch from the Sieve maintainer that adds the options
4449 sieve_useraddress and sieve_subaddress to the redirect router.
4451 PH/30 In these circumstances:
4452 . Two addresses routed to the same list of hosts;
4453 . First host does not offer TLS;
4454 . First host accepts first address;
4455 . First host gives temporary error to second address;
4456 . Second host offers TLS and a TLS session is established;
4457 . Second host accepts second address.
4458 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4459 and peerdn, if requested) that were in fact used only for the second
4462 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4463 attention to any local part prefix or suffix that was matched by the
4464 router that accepted the address. It now behaves in the same way as it
4465 does for delivery: the affixes are removed from the local part unless
4466 rcpt_include_affixes is set on the transport.
4468 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4469 timeout during the DATA phase of an incoming message.
4471 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4472 applied a patch sanctioned by the Sieve maintainer.
4474 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4475 the uid or gid is negative. A case of a negative gid caused this to be
4476 noticed. The fix allows for either to be negative.
4478 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4479 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4482 PH/36 The change PH/12 above was broken. Fixed it.
4484 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4485 the grounds that routing the same address twice would always produce the
4486 same answer. This might have been true once, but it is certainly no
4487 longer true now. Routing a child address may depend on the previous
4488 routing that produced that child. Some complicated redirection strategies
4489 went wrong when messages had multiple recipients, and made Exim's
4490 behaviour dependent on the order in which the addresses were given.
4492 I have moved the duplicate checking until after the routing is complete.
4493 Exim scans the addresses that are assigned to local and remote
4494 transports, and removes any duplicates. This means that more work will be
4495 done, as duplicates will always all be routed, but duplicates are
4496 presumably rare, so I don't expect this is of any significance.
4498 For deliveries to pipes, files, and autoreplies, the duplicate checking
4499 still happens during the routing process, since they are not going to be
4502 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4503 It corrects a timeout issue with spamd. This is Ian's comment: "The
4504 background is that sometimes spamd either never reads data from a
4505 connection it has accepted, or it never writes response data. The exiscan
4506 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4507 blindly assumes that writes won't block so it may never time out."
4509 PH/39 Allow G after quota size as well as K and M.
4511 PH/40 The value set for $authenticated_id in an authenticator may not contain
4512 binary zeroes or newlines because the value is written to log lines and
4513 to spool files. There was no check on this. Now the value is run through
4514 the string_printing() function so that such characters are converted to
4515 printable escape sequences.
4517 PH/41 $message_linecount is a new variable that contains the total number of
4518 lines in the message. Compare $body_linecount, which is the count for the
4521 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4522 bounce and delay warning messages, except in certain special cases, which
4525 (a) An SMTP error message from a remote host;
4526 (b) A message specified in a :fail: redirection;
4527 (c) A message specified in a "fail" command in a system filter;
4528 (d) A message specified in a FAIL return from the queryprogram router;
4529 (e) A message specified by the cannot_route_message router option.
4531 In these cases only, Exim does include the error details in bounce and
4532 warning messages. There are also a few cases where bland messages such
4533 as "unrouteable address" or "local delivery error" are given.
4535 PH/43 $value is now also set for the "else" part of a ${run expansion.
4537 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4538 being worked on, but at least Exim now implements the latest version to
4541 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4542 process to complete was treated as a delivery failure, a timeout while
4543 writing the message to the pipe was logged, but erroneously treated as a
4544 successful delivery. Such timeouts include transport filter timeouts. For
4545 consistency with the overall process timeout, these timeouts are now
4546 treated as errors, giving rise to delivery failures by default. However,
4547 there is now a new Boolean option for the pipe transport called
4548 timeout_defer, which, if set TRUE, converts the failures into defers for
4549 both kinds of timeout. A transport filter timeout is now identified in
4552 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4553 systems where "make" and "gmake" are different, calling "gmake" at top
4554 level broke things. I've arranged for the value of $(MAKE) to be passed
4555 from the Makefile to this script so that it can call the same version of
4559 A note about Exim versions 4.44 and 4.50
4560 ----------------------------------------
4562 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4563 changes of various kinds. As a consequence, a big documentation update was
4564 needed. This delayed the release for rather longer than seemed good, especially
4565 in the light of a couple of (minor) security issues. Therefore, the changes
4566 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4567 release. So 4.44 and 4.50 are in effect two different branches that both start
4570 I have left the 4.50 change log unchanged; it contains all the changes since
4571 4.43. The change log for 4.44 is below; many of its items are identical to
4572 those for 4.50. This seems to be the most sensible way to preserve the
4573 historical information.
4579 1. Minor wording change to the doc/README.SIEVE file.
4581 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4582 computation of the current number of files was incorrect.
4584 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4585 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4588 4. Give more explanation in the error message when the command for a transport
4589 filter fails to execute.
4591 5. There are several places where Exim runs a non-Exim command in a
4592 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4593 was being done only for the command run by the queryprogram router. It is
4594 now done for all such subprocesses. The other cases are: ${run, transport
4595 filters, and the commands run by the lmtp and pipe transports.
4597 6. Added CONFIGURE_GROUP build-time option.
4599 7. Some older OS have a limit of 256 on the maximum number of file
4600 descriptors. Exim was using setrlimit() to set 1000 as a large value
4601 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4602 systems. I've change it so that if it can't get 1000, it tries for 256.
4604 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4605 was an oversight, and furthermore, ever since the addition of extra
4606 controls (e.g. 4.43/32), the checks on when to allow different forms of
4607 "control" were broken. There should now be diagnostics for all cases when a
4608 control that does not make sense is encountered.
4610 9. Added the /retain_sender option to "control=submission".
4612 10. $recipients is now available in the predata ACL (oversight).
4614 11. Tidy the search cache before the fork to do a delivery from a message
4615 received from the command line. Otherwise the child will trigger a lookup
4616 failure and thereby defer the delivery if it tries to use (for example) a
4617 cached ldap connection that the parent has called unbind on.
4619 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4620 of $address_data from the recipient verification was clobbered by the
4621 sender verification.
4623 13. The value of address_data from a sender verification is now available in
4624 $sender_address_data in subsequent conditions in the ACL statement.
4626 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4628 15. Added a new option "connect=<time>" to callout options, to set a different
4631 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4632 was its contents. (It was OK if the option was not defined at all.)
4634 17. A "Completed" log line is now written for messages that are removed from
4635 the spool by the -Mrm option.
4637 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4638 information about exactly what failed.
4640 19. Added -dd to debug only the daemon process.
4642 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4643 handles timeouts, both on the server side and network timeouts. Renamed the
4644 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4646 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4647 It is now set to "smtps".
4649 22. $host_address is now set to the target address during the checking of
4650 ignore_target_hosts.
4652 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4653 being passed; this would have caused $sender_host_name to have been used if
4654 matching the list had actually called for a host name (not very likely,
4655 since this list is usually IP addresses). A host name is now passed as
4658 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4659 code that specifies a non-restarting handler (typically sigaction() in
4660 modern systems) in an attempt to fix a rare and obscure crash bug.
4662 25. Narrowed the window for a race in the daemon that could cause it to ignore
4663 SIGCHLD signals. This is not a major problem, because they are used only to
4664 wake it up if nothing else does.
4666 26. A malformed maildirsize file could cause Exim to calculate negative values
4667 for the mailbox size or file count. Odd effects could occur as a result.
4668 The maildirsize information is now recalculated if the size or filecount
4671 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4672 support for a long time. Removed HAVE_SYS_VFS_H.
4674 28. Installed the latest version of exipick from John Jetmore.
4676 29. In an address list, if the pattern was not a regular expression, an empty
4677 subject address (from a bounce message) matched only if the pattern was an
4678 empty string. Non-empty patterns were not even tested. This was the wrong
4679 because it is perfectly reasonable to use an empty address as part of a
4680 database query. An empty address is now tested by patterns that are
4681 lookups. However, all the other forms of pattern expect the subject to
4682 contain a local part and a domain, and therefore, for them, an empty
4683 address still always fails if the pattern is not itself empty.
4685 30. Exim went into a mad DNS loop when attempting to do a callout where the
4686 host was specified on an smtp transport, and looking it up yielded more
4687 than one IP address.
4689 31. Re-factored the code for checking spool and log partition space into a
4690 function that finds that data and another that does the check. The former
4691 is then used to implement four new variables: $spool_space, $log_space,
4692 $spool_inodes, and $log_inodes.
4694 32. The RFC2047 encoding function was originally intended for short strings
4695 such as real names; it was not keeping to the 75-character limit for
4696 encoded words that the RFC imposes. It now respects the limit, and
4697 generates multiple encoded words if necessary. To be on the safe side, I
4698 have increased the buffer size for the ${rfc2047: expansion operator from
4701 33. It is now permitted to omit both strings after an "if" condition; if the
4702 condition is true, the result is "true". As before, when the second string
4703 is omitted, a false condition yields an empty string. This makes it less
4704 cumbersome to write custom ACL and router conditions.
4706 34. Failure to deliver a bounce message always caused it to be frozen, even if
4707 there was an errors_to setting on the router. The errors_to setting is now
4710 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4711 canonical form (fully expanded) before being placed in
4712 $sender_host_address.
4714 36. The table in the code that translates DNS record types into text (T_A to
4715 "A" for instance) was missing entries for NS and CNAME. It is just possible
4716 that this could have caused confusion if both these types were looked up
4717 for the same domain, because the text type is used as part of Exim's
4718 per-process caching. But the chance of anyone hitting this buglet seems
4721 37. The dnsdb lookup has been extended in a number of ways.
4723 (1) There is a new type, "zns", which walks up the domain tree until it
4724 finds some nameserver records. It should be used with care.
4726 (2) There is a new type, "mxh", which is like "mx" except that it returns
4727 just the host names, not the priorities.
4729 (3) It is now possible to give a list of domains (or IP addresses) to be
4730 looked up. The behaviour when one of the lookups defers can be
4731 controlled by a keyword.
4733 (4) It is now possible to specify the separator character for use when
4734 multiple records are returned.
4736 38. The dnslists ACL condition has been extended: it is now possible to supply
4737 a list of IP addresses and/or domains to be looked up in a particular DNS
4740 39. Added log_selector=+queue_time_overall.
4742 40. When running the queue in the test harness, wait just a tad after forking a
4743 delivery process, to get repeatability of debugging output.
4745 41. Include certificate and key file names in error message when GnuTLS fails
4746 to set them up, because the GnuTLS error message doesn't include the name
4747 of the failing file when there is a problem reading it.
4749 42. Allow both -bf and -bF in the same test run.
4751 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4753 44. The "Exiscan patch" is now merged into the mainline Exim source.
4755 45. Sometimes the final signoff response after QUIT could fail to get
4756 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4757 < 0 before doing a fflush(). This bug looks as though it goes back to the
4758 introduction of TLS in release 3.20, but "sometimes" must have been rare
4759 because the tests only now provoked it.
4761 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4762 (this can affect the format of dates).
4764 47. exim_tidydb, when checking for the continued existence of a message for
4765 which it has found a message-specific retry record, was not finding
4766 messages that were in split spool directories. Consequently, it was
4767 deleting retry records that should have stayed in existence.
4769 48. Steve fixed some bugs in eximstats.
4771 49. The SPA authentication driver was not abandoning authentication and moving
4772 on to the next authenticator when an expansion was forced to fail,
4773 contradicting the general specification for all authenticators. Instead it
4774 was generating a temporary error. It now behaves as specified.
4776 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4777 (the order specifies the preference for clients). The order is now AES256,
4778 AES128, 3DES, ARCFOUR128.
4780 51. Small patch to Sieve code - explicitly set From: when generating an
4783 52. Exim crashed if a remote delivery caused a very long error message to be
4784 recorded - for instance if somebody sent an entire SpamAssassin report back
4785 as a large number of 550 error lines. This bug was coincidentally fixed by
4786 increasing the size of one of Exim's internal buffers (big_buffer) that
4787 happened as part of the Exiscan merge. However, to be on the safe side, I
4788 have made the code more robust (and fixed the comments that describe what
4791 53. Now that there can be additional text after "Completed" in log lines (if
4792 the queue_time_overall log selector is set), a one-byte patch to exigrep
4793 was needed to allow it to recognize "Completed" as not the last thing in
4796 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4797 patch that reportedly fixes this has been added. I am not expert enough to
4798 create a test for it. This is what the patch creator wrote:
4800 "I found a little strange behaviour of ldap code when working with
4801 Windows 2003 AD Domain, where users was placed in more than one
4802 Organization Units. When I tried to give exim partial DN, the exit code
4803 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4804 But simultaneously result of request was absolutely normal ldap result,
4805 so I produce this patch..."
4807 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4808 REFERENCE, so I have modified the code to exclude the patch when that macro
4811 55. Some experimental protocols are using DNS PTR records for new purposes. The
4812 keys for these records are domain names, not reversed IP addresses. The
4813 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4814 leaves it alone. Component reversal etc. now happens only for IP addresses.
4817 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4819 57. Double the size of the debug message buffer (to 2048) so that more of very
4820 long debug lines gets shown.
4822 58. The exicyclog utility now does better if the number of log files to keep
4823 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4825 59. Two changes related to the smtp_active_hostname option:
4827 (1) $smtp_active_hostname is now available as a variable.
4828 (2) The default for smtp_banner uses $smtp_active_hostname instead
4829 of $primary_hostname.
4831 60. The host_aton() function is supposed to be passed a string that is known
4832 to be a valid IP address. However, in the case of IPv6 addresses, it was
4833 not checking this. This is a hostage to fortune. Exim now panics and dies
4834 if the condition is not met. A case was found where this could be provoked
4835 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4836 components; fortuitously, this particular loophole had already been fixed
4837 by change 4.50/55 above.
4839 If there are any other similar loopholes, the new check in host_aton()
4840 itself should stop them being exploited. The report I received stated that
4841 data on the command line could provoke the exploit when Exim was running as
4842 exim, but did not say which command line option was involved. All I could
4843 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4844 running as the user.
4847 61. There was a buffer overflow vulnerability in the SPA authentication code
4848 (which came originally from the Samba project). I have added a test to the
4849 spa_base64_to_bits() function which I hope fixes it.
4852 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4853 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4855 63. The daemon start-up calls getloadavg() while still root for those OS that
4856 need the first call to be done as root, but it missed one case: when
4857 deliver_queue_load_max is set with deliver_drop_privilege. This is
4858 necessary for the benefit of the queue runner, because there is no re-exec
4859 when deliver_drop_privilege is set.
4861 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4862 This has been fixed.
4864 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4865 was in use, was not putting the data itself into the right store pool;
4866 consequently, it could be overwritten for a subsequent message in the same
4867 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4870 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4872 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4873 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4874 the function were treating the return as a boolean value, which happened to
4875 work because 0=false and not-0=true, but is not correct code.
4877 68. The host_aton() function was not handling scoped IPv6 addresses (those
4878 with, for example, "%eth0" on the end) correctly.
4880 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4881 negated items (that is, ~something) in unsigned ints. Some compilers
4882 apparently mutter when there is no cast.
4884 70. If an address verification called from an ACL failed, and did not produce a
4885 user-specific message (i.e. there was only a "system" message), nothing was
4886 put in $acl_verify_message. In this situation, it now puts the system
4889 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4890 session; change 4.31/43 added the unwanted input to the log line - except
4891 that it did not do this in the start of session case. It now does.
4893 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4894 This is wrong and can cause the other end to generate a synchronization
4895 error if it is another Exim or anything else that does the synchronization
4896 check. A QUIT command is no longer sent after a timeout.
4898 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4899 during host lookups.
4901 74. The defer_ok option of callout verification was not working if it was used
4902 when verifying addresses in header lines, that is, for this case:
4904 verify = header_sender/callout=defer_ok
4906 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4907 those file descriptors could be used for SMTP connections. If anything
4908 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4909 could be sent to the SMTP client, causing chaos. The daemon now opens
4910 stdin, stdout, and stderr to /dev/null when it puts itself into the
4913 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4914 log by default. The user can override this with suitable Perl magic.
4916 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4917 the log message when discard triggers, was not working for the DATA ACL or
4918 for the non-SMTP ACL.
4920 78. Error message wording change in sieve.c.
4922 79. If smtp_accept_max_per_host was set, the number of connections could be
4923 restricted to fewer than expected, because the daemon was trying to set up
4924 a new connection before checking whether the processes handling previous
4925 connections had finished. The check for completed processes is now done
4926 earlier. On busy systems, this bug wouldn't be noticed because something
4927 else would have woken the daemon, and it would have reaped the completed
4930 80. If a message was submitted locally by a user whose login name contained one
4931 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4932 It caused a spool format error. I have fixed the spool reading code. A
4933 related problem was that the "from" clause in the Received: line became
4934 illegal because of the space(s). It is now covered by ${quote_local_part.
4936 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4939 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4941 83. Patch from Sieve maintainer to fix unterminated string problem in
4942 "vacation" handling.
4944 84. Some minor changes to the Linux configuration files to help with other
4945 OS variants using glibc.
4947 85. One more patch for Sieve to update vacation handling to latest spec.
4950 ----------------------------------------------------
4951 See the note above about the 4.44 and 4.50 releases.
4952 ----------------------------------------------------
4958 1. Change 4.43/35 introduced a bug that caused file counts to be
4959 incorrectly computed when quota_filecount was set in an appendfile
4962 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4963 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4966 3. Give more explanation in the error message when the command for a transport
4967 filter fails to execute.
4969 4. There are several places where Exim runs a non-Exim command in a
4970 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4971 was being done only for the command run by the queryprogram router. It is
4972 now done for all such subprocesses. The other cases are: ${run, transport
4973 filters, and the commands run by the lmtp and pipe transports.
4975 5. Some older OS have a limit of 256 on the maximum number of file
4976 descriptors. Exim was using setrlimit() to set 1000 as a large value
4977 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4978 systems. I've change it so that if it can't get 1000, it tries for 256.
4980 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4981 was an oversight, and furthermore, ever since the addition of extra
4982 controls (e.g. 4.43/32), the checks on when to allow different forms of
4983 "control" were broken. There should now be diagnostics for all cases when a
4984 control that does not make sense is encountered.
4986 7. $recipients is now available in the predata ACL (oversight).
4988 8. Tidy the search cache before the fork to do a delivery from a message
4989 received from the command line. Otherwise the child will trigger a lookup
4990 failure and thereby defer the delivery if it tries to use (for example) a
4991 cached ldap connection that the parent has called unbind on.
4993 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4994 of $address_data from the recipient verification was clobbered by the
4995 sender verification.
4997 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4998 was its contents. (It was OK if the option was not defined at all.)
5000 11. A "Completed" log line is now written for messages that are removed from
5001 the spool by the -Mrm option.
5003 12. $host_address is now set to the target address during the checking of
5004 ignore_target_hosts.
5006 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5007 being passed; this would have caused $sender_host_name to have been used if
5008 matching the list had actually called for a host name (not very likely,
5009 since this list is usually IP addresses). A host name is now passed as
5012 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5013 code that specifies a non-restarting handler (typically sigaction() in
5014 modern systems) in an attempt to fix a rare and obscure crash bug.
5016 15. Narrowed the window for a race in the daemon that could cause it to ignore
5017 SIGCHLD signals. This is not a major problem, because they are used only to
5018 wake it up if nothing else does.
5020 16. A malformed maildirsize file could cause Exim to calculate negative values
5021 for the mailbox size or file count. Odd effects could occur as a result.
5022 The maildirsize information is now recalculated if the size or filecount
5025 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5026 support for a long time. Removed HAVE_SYS_VFS_H.
5028 18. Updated exipick to current release from John Jetmore.
5030 19. Allow an empty sender to be matched against a lookup in an address list.
5031 Previously the only cases considered were a regular expression, or an
5034 20. Exim went into a mad DNS lookup loop when doing a callout where the
5035 host was specified on the transport, if the DNS lookup yielded more than
5038 21. The RFC2047 encoding function was originally intended for short strings
5039 such as real names; it was not keeping to the 75-character limit for
5040 encoded words that the RFC imposes. It now respects the limit, and
5041 generates multiple encoded words if necessary. To be on the safe side, I
5042 have increased the buffer size for the ${rfc2047: expansion operator from
5045 22. Failure to deliver a bounce message always caused it to be frozen, even if
5046 there was an errors_to setting on the router. The errors_to setting is now
5049 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5050 canonical form (fully expanded) before being placed in
5051 $sender_host_address.
5053 24. Updated eximstats to version 1.33
5055 25. Include certificate and key file names in error message when GnuTLS fails
5056 to set them up, because the GnuTLS error message doesn't include the name
5057 of the failing file when there is a problem reading it.
5059 26. Expand error message when OpenSSL has problems setting up cert/key files.
5062 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5063 (this can affect the format of dates).
5065 28. exim_tidydb, when checking for the continued existence of a message for
5066 which it has found a message-specific retry record, was not finding
5067 messages that were in split spool directories. Consequently, it was
5068 deleting retry records that should have stayed in existence.
5070 29. eximstats updated to version 1.35
5071 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5072 1.35 - bugfix such that pie charts by volume are generated correctly
5074 30. The SPA authentication driver was not abandoning authentication and moving
5075 on to the next authenticator when an expansion was forced to fail,
5076 contradicting the general specification for all authenticators. Instead it
5077 was generating a temporary error. It now behaves as specified.
5079 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5080 (the order specifies the preference for clients). The order is now AES256,
5081 AES128, 3DES, ARCFOUR128.
5083 31. Small patch to Sieve code - explicitly set From: when generating an
5086 32. Exim crashed if a remote delivery caused a very long error message to be
5087 recorded - for instance if somebody sent an entire SpamAssassin report back
5088 as a large number of 550 error lines. This bug was coincidentally fixed by
5089 increasing the size of one of Exim's internal buffers (big_buffer) that
5090 happened as part of the Exiscan merge. However, to be on the safe side, I
5091 have made the code more robust (and fixed the comments that describe what
5094 33. Some experimental protocols are using DNS PTR records for new purposes. The
5095 keys for these records are domain names, not reversed IP addresses. The
5096 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5097 leaves it alone. Component reversal etc. now happens only for IP addresses.
5100 34. The host_aton() function is supposed to be passed a string that is known
5101 to be a valid IP address. However, in the case of IPv6 addresses, it was
5102 not checking this. This is a hostage to fortune. Exim now panics and dies
5103 if the condition is not met. A case was found where this could be provoked
5104 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5105 components; fortuitously, this particular loophole had already been fixed
5106 by change 4.50/55 or 4.44/33 above.
5108 If there are any other similar loopholes, the new check in host_aton()
5109 itself should stop them being exploited. The report I received stated that
5110 data on the command line could provoke the exploit when Exim was running as
5111 exim, but did not say which command line option was involved. All I could
5112 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5113 running as the user.
5116 35. There was a buffer overflow vulnerability in the SPA authentication code
5117 (which came originally from the Samba project). I have added a test to the
5118 spa_base64_to_bits() function which I hope fixes it.
5121 36. The daemon start-up calls getloadavg() while still root for those OS that
5122 need the first call to be done as root, but it missed one case: when
5123 deliver_queue_load_max is set with deliver_drop_privilege. This is
5124 necessary for the benefit of the queue runner, because there is no re-exec
5125 when deliver_drop_privilege is set.
5127 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5128 was in use, was not putting the data itself into the right store pool;
5129 consequently, it could be overwritten for a subsequent message in the same
5130 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5133 38. Sometimes the final signoff response after QUIT could fail to get
5134 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5135 < 0 before doing a fflush(). This bug looks as though it goes back to the
5136 introduction of TLS in release 3.20, but "sometimes" must have been rare
5137 because the tests only now provoked it.
5143 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5144 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5145 Now it returns an integer. A number of calls were still expecting a T/F
5146 return. Fortuitously, in all cases, the tests worked in OK situations,
5147 which is the norm. However, things would have gone wrong on any write
5148 failures on the smtp file descriptor. This function is used when sending
5149 messages over SMTP and also when doing verify callouts.
5151 2. When Exim is called to do synchronous delivery of a locally submitted
5152 message (the -odf or -odi options), it no longer closes stderr before doing
5155 3. Implemented the mua_wrapper option.
5157 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5159 5. Implemented the functions header_remove(), header_testname(),
5160 header_add_at_position(), and receive_remove_recipient(), and exported them
5163 6. If an ACL "warn" statement specified the addition of headers, Exim already
5164 inserted X-ACL-Warn: at the start if there was no header name. However, it
5165 was not making this test for the second and subsequent header lines if
5166 there were newlines in the string. This meant that an invalid header could
5167 be inserted if Exim was badly configured.
5169 7. Allow an ACL "warn" statement to add header lines at the start or after all
5170 the Received: headers, as well as at the end.
5172 8. Added the rcpt_4xx retry error code.
5174 9. Added postmaster_mailfrom=xxx to callout verification option.
5176 10. Added mailfrom=xxxx to the callout verification option, for verify=
5179 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5180 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5182 12. Inserted some casts to stop certain compilers warning when using pointer
5183 differences as field lengths or precisions in printf-type calls (mostly
5184 affecting debugging statements).
5186 13. Added optional readline() support for -be (dynamically loaded).
5188 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5189 same clock tick as a message's arrival, so that its received time was the
5190 same as the "first fail" time on the retry record, and that message
5191 remained on the queue past the ultimate address timeout, every queue runner
5192 would try a delivery (because it was past the ultimate address timeout) but
5193 after another failure, the ultimate address timeout, which should have then
5194 bounced the address, did not kick in. This was a "< instead of <=" error;
5195 in most cases the first failure would have been in the next clock tick
5196 after the received time, and all would be well.
5198 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5199 being recognized when the domain list was tested by the match_domain
5200 condition in an expansion string.
5202 16. Added the ${str2b64: operator.
5204 17. Exim was always calling setrlimit() to set a large limit for the number of
5205 processes, without checking whether the existing limit was already
5206 adequate. (It did check for the limit on file descriptors.) Furthermore,
5207 errors from getrlimit() and setrlimit() were being ignored. Now they are
5208 logged to the main and panic logs, but Exim does carry on, to try to do its
5209 job under whatever limits there are.
5211 18. Imported PCRE 5.0.
5213 19. Trivial typo in log message " temporarily refused connection" (the leading
5216 20. If the log selector return_path_on_delivery was set and an address was
5217 redirected to /dev/null, the delivery process crashed because it assumed
5218 that a return path would always be set for a "successful" delivery. In this
5219 case, the whole delivery is bypassed as an optimization, and therefore no
5222 21. Internal re-arrangement: the function for sending a challenge and reading
5223 a response while authentication was assuming a zero-terminated challenge
5224 string. It's now changed to take a pointer and a length, to allow for
5225 binary data in such strings.
5227 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5229 23. Exim was not respecting finduser_retries when seeking the login of the
5230 uid under which it was called; it was always trying 10 times. (The default
5231 setting of finduser_retries is zero.) Also, it was sleeping after the final
5232 failure, which is pointless.
5234 24. Implemented tls_on_connect_ports.
5236 25. Implemented acl_smtp_predata.
5238 26. If the domain in control=submission is set empty, Exim assumes that the
5239 authenticated id is a complete email address when it generates From: or
5240 Sender: header lines.
5242 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5243 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5244 chown and chgrp in /bin and hostname in /usr/bin.
5246 28. Exim was keeping the "process log" file open after each use, just as it
5247 does for the main log. This opens the possibility of it remaining open for
5248 long periods when the USR1 signal hits a daemon. Occasional processlog
5249 errors were reported, that could have been caused by this. Anyway, it seems
5250 much more sensible not to leave this file open at all, so that is what now
5253 29. The long-running daemon process does not normally write to the log once it
5254 has entered its main loop, and it closes the log before doing so. This is
5255 so that log files can straightforwardly be renamed and moved. However,
5256 there are a couple of unusual error situations where the daemon does write
5257 log entries, and I had neglected to close the log afterwards.
5259 30. The text of an SMTP error response that was received during a remote
5260 delivery was being truncated at 512 bytes. This is too short for some of
5261 the long messages that one sometimes sees. I've increased the limit to
5264 31. It is now possible to make retry rules that apply only when a message has a
5265 specific sender, in particular, an empty sender.
5267 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5268 it possible to be selective about when SMTP synchronization is enforced.
5270 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5272 32. Implemented hosts_connection_nolog.
5274 33. Added an ACL for QUIT.
5276 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5279 35. Added mailbox_size and mailbox_filecount to appendfile.
5281 36. Added control = no_multiline_responses to ACLs.
5283 37. There was a bug in the logic of the code that waits for the clock to tick
5284 in the case where the clock went backwards by a substantial amount such
5285 that the microsecond fraction of "now" was more than the microsecond
5286 fraction of "then" (but the whole seconds number was less).
5288 38. Added support for the libradius Radius client library this is found on
5289 FreeBSD (previously only the radiusclient library was supported).
5295 1. When certain lookups returned multiple values in the form name=value, the
5296 quoting of the values was not always being done properly. Specifically:
5297 (a) If the value started with a double quote, but contained no whitespace,
5299 (b) If the value contained whitespace other than a space character (i.e.
5300 tabs or newlines or carriage returns) it was not quoted.
5301 This fix has been applied to the mysql and pgsql lookups by writing a
5302 separate quoting function and calling it from the lookup code. The fix
5303 should probably also be applied to nisplus, ibase and oracle lookups, but
5304 since I cannot test any of those, I have not disturbed their existing code.
5306 2. A hit in the callout cache for a specific address caused a log line with no
5307 reason for rejecting RCPT. Now it says "Previous (cached) callout
5308 verification failure".
5310 3. There was an off-by-one bug in the queryprogram router. An over-long
5311 return line was truncated at 256 instead of 255 characters, thereby
5312 overflowing its buffer with the terminating zero. As well as fixing this, I
5313 have increased the buffer size to 1024 (and made a note to document this).
5315 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5316 when Exim is waiting for an SMTP response from a remote server, Exim
5317 restarts its select() call on the socket, thereby resetting its timeout.
5318 This is not a problem when such interrupts are rare. Somebody set up a cron
5319 job to run exiwhat every 2 minutes, which is less than the normal select()
5320 timeout (5 or 10 minutes). This meant that the select() timeout never
5321 kicked in because it was always reset. I have fixed this by comparing the
5322 time when an interrupt arrives with the time at the start of the first call
5323 to select(). If more time than the timeout has elapsed, the interrupt is
5324 treated as a timeout.
5326 5. Some internal re-factoring in preparation for the addition of Sieve
5327 extensions (by MH). In particular, the "personal" test is moved to a
5328 separate function, and given an option for scanning Cc: and Bcc: (which is
5329 not set for Exim filters).
5331 6. When Exim created an email address using the login of the caller as the
5332 local part (e.g. when creating a From: or Sender: header line), it was not
5333 quoting the local part when it contained special characters such as @.
5335 7. Installed new OpenBSD configuration files.
5337 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5338 try to make them clearer.
5340 9. Callout options, other than the timeout value, were being ignored when
5341 verifying sender addresses in header lines. For example, when using
5343 verify = header_sender/callout=no_cache
5345 the cache was (incorrectly) being used.
5347 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5348 only the Cygwin environment.
5350 11. When return_path_on_delivery was set as a log selector, if different remote
5351 addresses in the same message used different return paths and parallel
5352 remote delivery occurred, the wrong values would sometimes be logged.
5353 (Whenever a remote delivery process finished, the return path value from
5354 the most recently started remote delivery process was logged.)
5356 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5357 lines when AUTH and/or TLS are in use. This is the "received protocol"
5358 field. Exim used to use "asmtp" for authenticated SMTP, without any
5359 indication (in the protocol name) for TLS use. Now it follows the RFC and
5360 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5361 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5362 names appear in log lines as well as in Received: header lines.
5364 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5365 extensions, and comparison tests, and to fix some bugs.
5367 14. Changes to the "personal" filter test:
5369 (1) The test was buggy in that it was just doing the equivalent of
5370 "contains" tests on header lines. For example, if a user's address was
5371 anne@some.where, the "personal" test would incorrectly be true for
5373 To: susanne@some.where
5375 This test is now done by extracting each address from the header in turn,
5376 and checking the entire address. Other tests that are part of "personal"
5377 are now done using regular expressions (for example, to check local parts
5378 of addresses in From: header lines).
5380 (2) The list of non-personal local parts in From: addresses has been
5381 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5382 taken from the Sieve specification recommendations.
5384 (3) If the message contains any header line starting with "List-" it is
5385 treated as non-personal.
5387 (4) The test for "circular" in the Subject: header line has been removed
5388 because it now seems ill-conceived.
5390 15. Minor typos in src/EDITME comments corrected.
5392 16. Installed latest exipick from John Jetmore.
5394 17. If headers_add on a router specified a text string that was too long for
5395 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5396 of string_sprintf() is now avoided.
5398 18. $message_body_size was not set (it was always zero) when running the DATA
5399 ACL and the local_scan() function.
5401 19. For the "mail" command in an Exim filter, no default was being set for
5402 the once_repeat time, causing a random time value to be used if "once" was
5403 specified. (If the value happened to be <= 0, no repeat happened.) The
5404 default is now 0s, meaning "never repeat". The "vacation" command was OK
5405 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5406 (I found it when inspecting the code).
5408 20. There is now an overall timeout for performing a callout verification. It
5409 defaults to 4 times the callout timeout, which applies to individual SMTP
5410 commands during the callout. The overall timeout applies when there is more
5411 than one host that can be tried. The timeout is checked before trying the
5412 next host. This prevents very long delays if there are a large number of
5413 hosts and all are timing out (e.g. when the network connections are timing
5414 out). The value of the overall timeout can be changed by specifying an
5415 additional sub-option for "callout", called "maxwait". For example:
5417 verify = sender/callout=5s,maxwait=20s
5419 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5420 to the end before writing, but this should make it even safer).
5422 22. Exim was forgetting that it had advertised PIPELINING for the second and
5423 subsequent messages on an SMTP connection. It was also not resetting its
5424 memory on STARTTLS and an internal HELO.
5426 23. When Exim logs an SMTP synchronization error within a session, it now
5427 records whether PIPELINING has been advertised or not.
5429 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5430 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5431 rather than long int.
5433 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5435 26. Added the never_mail option to autoreply.
5441 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5442 crash if the getsockname() call failed; this can happen if a connection is
5443 closed very soon after it is established. The problem was simply in the
5444 order in which certain operations were done, causing Exim to try to write
5445 to the SMTP stream before it had set up the file descriptor. The bug has
5446 been fixed by making things happen in the correct order.
5452 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5453 before the connection was closed, thus losing the rejection response.
5455 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5456 some early Solaris releases, but causes trouble in current releases where
5457 socklen_t is defined.
5459 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5462 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5465 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5466 that releasing the top of it at the end releases what was used for sub-
5467 expansions (unless the block got too big). However, discard this block if
5468 the first thing is a variable or header, so that we can use its block when
5469 it is dynamic (useful for very large $message_headers, for example).
5471 6. Lookups now cache *every* query, not just the most recent. A new, separate
5472 store pool is used for this. It can be recovered when all lookup caches are
5473 flushed. Lookups now release memory at the end of their result strings.
5474 This has involved some general refactoring of the lookup sources.
5476 7. Some code has been added to the store_xxx() functions to reduce the amount
5477 of flapping under certain conditions.
5479 8. log_incoming_interface used to affect only the <= reception log lines. Now
5480 it causes the local interface and port to be added to several more SMTP log
5481 lines, for example "SMTP connection from", and rejection lines.
5483 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5485 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5487 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5488 could be overwritten at the end of the current message (or the start of a
5489 new message if it was set in a HELO ACL). The value is now preserved for
5490 the duration of the SMTP connection.
5492 12. If a transport had a headers_rewrite setting, and a matching header line
5493 contained an unqualified address, that address was qualified, even if it
5494 did not match any rewriting rules. The underlying bug was that the values
5495 of the flags that permit the existence of unqualified sender and recipient
5496 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5497 non-local messages, and by -bnq for local messages) were not being
5498 preserved with the message after it was received.
5500 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5501 "next input=" as part of the text comprising the host identity instead of
5502 the correct text. The code was using the same buffer for two different
5503 strings. However, depending on which order the printing function evaluated
5504 its arguments, the bug did not always show up. Under Linux, for example, my
5505 test suite worked just fine.
5507 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5508 doesn't work with some older versions of Perl. It has been changed to "my",
5509 which in any case is probably the better facility to use.
5511 15. A really picky compiler found some instances of statements for creating
5512 error messages that either had too many or two few arguments for the format
5515 16. The size of the buffer for calls to the DNS resolver has been increased
5516 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5517 for addresses that have a lot of PTR records. This alleviates a problem; it
5518 does not fully solve it.
5520 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5521 buffer now truncates the list and logs the incident, which is the same
5522 action as happens when Exim is looking up a host name and its aliases.
5523 Previously in this situation something unpredictable would happen;
5524 sometimes it was "internal error: store_reset failed".
5526 18. If a server dropped the connection unexpectedly when an Exim client was
5527 using GnuTLS and trying to read a response, the client delivery process
5528 crashed while trying to generate an error log message.
5530 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5531 string, for example:
5533 warn message = H1: something\nH2: something
5535 the text was added as a single header line from Exim's point of view
5536 though it ended up OK in the delivered message. However, searching for the
5537 second and subsequent header lines using $h_h2: did not work. This has been
5538 fixed. Similarly, if a system filter added multiple headers in this way,
5539 the routers could not see them.
5541 20. Expanded the error message when iplsearch is called with an invalid key to
5542 suggest using net-iplsearch in a host list.
5544 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5545 ACLs are no longer actually imposed (and a message to that effect is
5548 22. If a "gecos" field in a passwd entry contained escaped characters, in
5549 particular, if it contained a \" sequence, Exim got it wrong when building
5550 a From: or a Sender: header from that name. A second bug also caused
5551 incorrect handling when an unquoted " was present following a character
5552 that needed quoting.
5554 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5555 was not being matched caselessly.
5557 24. Arranged for all hyphens in the exim.8 source to be escaped with
5560 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5561 properly. Recipient callout cache records were still being keyed to include
5562 the sender, even when use_sender was set false. This led to far more
5563 callouts that were necessary. The sender is no longer included in the key
5564 when use_sender is false.
5566 26. Added "control = submission" modifier to ACLs.
5568 27. Added the ${base62d: operator to decode base 62 numbers.
5570 28. dnsdb lookups can now access SRV records.
5572 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5573 the configuration file.
5575 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5576 (-v) mode. This makes the output for a verbose queue run more intelligible.
5578 31. Added a use_postmaster feature to recipient callouts.
5580 32. Added the $body_zerocount variable, containing the number of binary zero
5581 bytes in the message body.
5583 33. The time of last modification of the "new" subdirectory is now used as the
5584 "mailbox time last read" when there is a quota error for a maildir
5587 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5589 35. Added +ignore_unknown as a special item in host lists.
5591 36. Code for decoding IPv6 addresses in host lists is now included, even if
5592 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5593 address was recognized as an IP address, but was then not correctly decoded
5594 into binary, causing unexpected and incorrect effects when compared with
5601 1. Very minor rewording of debugging text in manualroute to say "list of
5602 hosts" instead of "hostlist".
5604 2. If verify=header_syntax was set, and a header line with an unqualified
5605 address (no domain) and a large number of spaces between the end of the
5606 name and the colon was received, the reception process suffered a buffer
5607 overflow, and (when I tested it) crashed. This was caused by some obsolete
5608 code that should have been removed. The fix is to remove it!
5610 3. When running in the test harness, delay a bit after writing a bounce
5611 message to get a bit more predictability in the log output.
5613 4. Added a call to search_tidyup() just before forking a reception process. In
5614 theory, someone could use a lookup in the expansion of smtp_accept_max_
5615 per_host which, without the tidyup, could leave open a database connection.
5617 5. Added the variables $recipient_data and $sender_data which get set from a
5618 lookup success in an ACL "recipients" or "senders" condition, or a router
5619 "senders" option, similar to $domain_data and $local_part_data.
5621 6. Moved the writing of debug_print from before to after the "senders" test
5624 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5625 problems for message scanning, either using a data ACL, or using
5626 local_scan() because the Received: header was not generated till after they
5627 were called (in order to set the time as the time of reception completion).
5628 I have revised the way this works. The header is now generated after the
5629 body is received, but before the ACL or local_scan() are called. After they
5630 are run, the timestamp in the header is updated.
5636 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5637 before starting a queue runner without re-exec. This happened only when
5638 deliver_drop_privilege was set or when the Exim user was set to root. The
5639 effect of the bug was that timeouts during subsequent deliveries caused
5640 crashes instead of being properly handled. The handler is now left at its
5641 default (and expected) setting.
5643 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5644 message, again when deliver_drop_privilege is set or Exim is run as root.
5645 The bug described in (1) was not present in this case, but the tidying up
5646 of the other signals was missing. I have made the two cases consistent.
5648 3. The ignore_target_hosts setting on a manualroute router was being ignored
5649 for hosts that were looked up using the /MX notation.
5651 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5654 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5655 operated on the sender address. After changing the $sender_address to <>
5656 for the sender address verify, Exim was re-instated it as the original
5657 (before rewriting) address, but remembering that it had rewritten it, so it
5658 wasn't rewriting it again. This bug also had the effect of breaking the
5659 sender address verification caching when the sender address was rewritten.
5661 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5662 This has been changed so that if the ip literal address matches
5663 ignore_target_hosts, the router declines.
5665 7. Added expansion conditions match_domain, match_address, and match_local_
5666 part (NOT match_host).
5668 8. The placeholder for the Received: header didn't have a length field set.
5670 9. Added code to Exim itself and to exim_lock to test for a specific race
5671 condition that could lead to file corruption when using MBX delivery. The
5672 issue is with the lockfile that is created in /tmp. If this file is removed
5673 after a process has opened it but before that process has acquired a lock,
5674 there is the potential for a second process to recreate the file and also
5675 acquire a lock. This could lead to two Exim processes writing to the file
5676 at the same time. The added code performs the same test as UW imapd; it
5677 checks after acquiring the lock that its file descriptor still refers to
5678 the same named file.
5680 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5681 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5684 11. Added the smtp_active_hostname option. If used, this will typically be made
5685 to depend on the incoming interface address. Because $interface_address is
5686 not set up until the daemon has forked a reception process, error responses
5687 that can happen earlier (such as "too many connections") no longer contain
5690 12. If an expansion in a condition on a "warn" statement fails because a lookup
5691 defers, the "warn" statement is abandoned, and the next ACL statement is
5692 processed. Previously this caused the whole ACL to be aborted.
5694 13. Added the iplsearch lookup type.
5696 14. Added ident_timeout as a log selector.
5698 15. Added tls_certificate_verified as a log selector.
5700 16. Added a global option tls_require_ciphers (compare the smtp transport
5701 option of the same name). This controls incoming TLS connections.
5703 17. I finally figured out how to make tls_require_ciphers do a similar thing
5704 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5705 before starting the TLS session.
5707 18. Tabs are now shown as \t in -bP output.
5709 19. If the log selector return_path_on_delivery was set, Exim crashed when
5710 bouncing a message because it had too many Received: header lines.
5712 20. If two routers both had headers_remove settings, and the first one included
5713 a superfluous trailing colon, the final name in the first list and the
5714 first name in the second list were incorrectly joined into one item (with a
5715 colon in the middle).
5721 1. Added -C and -D options to the exinext utility, mainly to make it easier
5722 to include in the automated testing, but these could be helpful when
5723 multiple configurations are in use.
5725 2. The exinext utility was not formatting the output nicely when there was
5726 an alternate port involved in the retry record key, nor when there was a
5727 message id as well (for retries that were specific to a specific message
5728 and a specific host). It was also confused by IPv6 addresses, because of
5729 the additional colons they contain. I have fixed the IPv4 problem, and
5730 patched it up to do a reasonable job for IPv6.
5732 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5733 delivery, the log line now contains "pipelined" if PIPELINING was used.
5735 4. An SMTP transport process used to panic and die if the bind() call to set
5736 an explicit outgoing interface failed. This has been changed; it is now
5737 treated in the same way as a connect() failure.
5739 5. A reference to $sender_host_name in the part of a conditional expansion
5740 that was being skipped was still causing a DNS lookup. This no longer
5743 6. The def: expansion condition was not recognizing references to header lines
5744 that used bh_ and bheader_.
5746 7. Added the _cache feature to named lists.
5748 8. The code for checking quota_filecount in the appendfile transport was
5749 allowing one more file than it should have been.
5751 9. For compatibility with Sendmail, the command line option
5759 and sets the incoming protocol and host name (for trusted callers). The
5760 host name and its colon can be omitted when only the protocol is to be set.
5761 Note the Exim already has two private options, -pd and -ps, that refer to
5762 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5763 "s", but I don't think that's a major issue.
5765 10. A number of refactoring changes to the code, none of which should affect
5768 (a) The number of logging options was getting close to filling up the
5769 32-bit word that was used as a bit map. I have split them into two classes:
5770 those that are passed in the argument to log_write(), and those that are
5771 only ever tested independently outside of that function. These are now in
5772 separate 32-bit words, so there is plenty of room for expansion again.
5773 There is no change in the user interface or the logging behaviour.
5775 (b) When building, for example, log lines, the code previously used a
5776 macro that called string_cat() twice, in order to add two strings. This is
5777 not really sufficiently general. Furthermore, there was one instance where
5778 it was actually wrong because one of the argument was used twice, and in
5779 one call a function was used. (As it happened, calling the function twice
5780 did not affect the overall behaviour.) The macro has been replaced by a
5781 function that can join an arbitrary number of extra strings onto a growing
5784 (c) The code for expansion conditions now uses a table and a binary chop
5785 instead of a serial search (which was left over from when there were very
5786 few conditions). Also, it now recognizes conditions like "pam" even when
5787 the relevant support is not compiled in: a suitably worded error message is
5788 given if an attempt is made to use such a condition.
5790 11. Added ${time_interval:xxxxx}.
5792 12. A bug was causing one of the ddress fields not to be passed back correctly
5793 from remote delivery subprocesses. The field in question was not being
5794 subsequently used, so this caused to problems in practice.
5796 13. Added new log selectors queue_time and deliver_time.
5798 14. Might have fixed a bug in maildirsizefile handling that threw up
5799 "unexpected character" debug warnings, and recalculated the data
5800 unnecessarily. In any case, I expanded the warning message to give more
5803 15. Added the message "Restricted characters in address" to the statements in
5804 the default ACL that block characters like @ and % in local parts.
5806 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5807 Three changes have been made:
5809 (a) There was a serious bug; a negative response to MAIL caused the whole
5810 recipient domain to be cached as invalid, thereby blocking all messages
5811 to all local parts at the same domain, from all senders. This bug has
5812 been fixed. The domain is no longer cached after a negative response to
5813 MAIL if the sender used is not empty.
5815 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5818 (c) A new callout option, "use_sender" has been added for people who want
5819 the modified behaviour.
5825 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5828 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5829 indeed breaks things for older releases.
5831 3. Added additional logging to the case where there is a problem reading data
5832 from a filter that is running in a subprocess using a pipe, in order to
5833 try to track down a specific problem.
5835 4. Testing facility fudge: when running in the test harness and attempting
5836 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5837 getting "No route to host". Convert this to a timeout.
5839 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5842 6. Some OS don't have socklen_t but use size_t instead. This affects the
5843 fifth argument of getsockopt() amongst other things. This is now
5844 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5845 can be set for individual OS. I have set it for SunOS5, OSF1, and
5846 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5847 some earlier ones do not.
5849 7. Change 4.30/15 was not doing the test caselessly.
5851 8. The standard form for an IPv6 address literal was being rejected by address
5852 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5853 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5854 this, as well as the form without the "IPv6" on the front (but only when
5855 address literals are enabled, of course).
5857 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5859 10. Exim crashed if a message with an empty sender address specified by -f
5860 encountered a router with an errors_to setting. This could be provoked only
5861 by a command such as
5865 where an empty string was supplied; "<>" did not hit this bug.
5867 11. Installed PCRE release 4.5.
5869 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5870 remained set. It is now erased.
5872 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5873 times from message ids (which are base 36 rather than the normal 62).
5875 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5876 were being counted as actual protocol errors, and logged if the log
5877 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5878 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5879 and DATA following a set of rejected RCPTs do not count as protocol errors.
5880 In other words, Exim assumes they were pipelined, though this may not
5881 actually be the case. Of course, in all cases the client gets an
5882 appropriate error code.
5884 15. If a lookup fails in an ACL condition, a message about the failure may
5885 be available; it is used if testing the ACL cannot continue, because most
5886 such messages specify what the cause of the deferral is. However, some
5887 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5888 that caused an old message to be retained and used if a later statement
5889 caused a defer, replacing the real cause of the deferral.
5891 16. If an IP address had so many PTR records that the DNS lookup buffer
5892 was not large enough to hold them, Exim could crash while trying to process
5893 the truncated data. It now detects and logs this case.
5895 17. Further to 4.21/58, another change has been made: if (and only if) the
5896 first line of a message (the first header line) ends with CRLF, a bare LF
5897 in a subsequent header line has a space inserted after it, so as not to
5898 terminate the header.
5900 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5901 unnecessarily, used atoi() instead of strtol(), and didn't check the
5902 termination when getting file sizes from file names by regex.
5904 19. Completely re-implemented the support for maildirsize files, in the light
5905 of a number of problems with the previous contributed implementation
5906 (4.30/29). In particular:
5908 . If the quota is zero, the maildirsize file is maintained, but no quota is
5911 . If the maildir directory does not exist, it is created before any attempt
5912 to write a maildirsize file.
5914 . The quota value in the file is just a cache; if the quota is changed in
5915 the transport, the new value overrides.
5917 . A regular expression is available for excluding directories from the
5920 20. The autoreply transport checks the characters in options that define the
5921 message's headers; it allows continued headers, but it was checking with
5922 isspace() after an embedded newline instead of explicitly looking for a
5925 21. If all the "regular" hosts to which an address was routed had passed their
5926 expiry times, and had not reached their retry times, the address was
5927 bounced, even if fallback hosts were defined. Now Exim should go on to try
5930 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5931 equivalent code in the SMTP transport. Some hosts send humungous responses
5932 to HELO/EHLO, more than 1024 it seems.
5934 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5935 gives compiler warnings in some environments. I've now done it "properly",
5938 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5939 (because of problems with the built-in one) was declared to return uschar *
5940 instead of char *, causing compiler failure.
5942 25. Fixed a file descriptor leak when processing alias/forward files.
5944 26. Fixed a minor format string issue in dbfn.c.
5946 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5948 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5949 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5952 29. When a sender address is verified, it is cached, to save repeating the test
5953 when there is more than one recipient in a message. However, when the
5954 verification involves a callout, it is possible for different callout
5955 options to be set for different recipients. It is too complicated to keep
5956 track of this in the cache, so now Exim always runs a verification when a
5957 callout is required, relying on the callout cache for the optimization.
5958 The overhead is duplication of the address routing, but this should not be
5961 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5962 to be verified with callout=postmaster, and the main callout worked but the
5963 postmaster check failed, the verification correctly failed. However, if a
5964 subsequent RCPT command asked for sender verification *without* the
5965 postmaster check, incorrect caching caused this verification also to fail,
5968 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5969 it was not caching the DNS options (qualify_single, search_parents) that
5970 were used when the lookup failed. A subsequent lookup with different
5971 options therefore always gave the same answer, though there were cases
5972 where it should not have. (Example: a "domains = !$mx_any" option on a
5973 dnslookup router: the "domains" option is always processed without any
5974 widening, but the router might have qualify_single set.) Now Exim uses the
5975 cached value only when the same options are set.
5977 32. Added John Jetmore's "exipick" utility to the distribution.
5979 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5980 than a timeout (e.g. a certificate is required, and is not provided), an
5981 Exim server now closes the connection immediately. Previously it waited for
5982 the client to close - but if the client is SSL, it seems that they each
5983 wait for each other, leading to a delay before one of them times out.
5985 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5986 maintained 0.8.x compatibility because I don't think many are using it, and
5987 it is clearly obsolete.
5989 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5992 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5993 client certificate was expired. A simple patch fixes this, though I don't
5994 understand the full logic of why the verify callback is called multiple
5997 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5998 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5999 which causes problems with some clients (such as the Certicom SSL Plus
6000 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6001 disables the coutermeasure allowing Eudora to connect."
6003 38. Exim was not checking that a write() to a log file succeeded. This could
6004 lead to Bad Things if a log got too big, in particular if it hit a file
6005 size limit. Exim now panics and dies if it cannot write to a log file, just
6006 as it does if it cannot open a log file.
6008 39. Modified OS/Makefile-Linux so that it now contains
6010 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6012 The two -D definitions ensure that Exim is compiled with large file
6013 support, which makes it possible to handle log files that are bigger than
6016 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6017 instance) a domain was checked against a named list that involved a lookup,
6018 causing $domain_data to be set, then another domain was checked against the
6019 same list, then the first domain was re-checked, the value of $domain_data
6020 after the final check could be wrong. In particular, if the second check
6021 failed, it could be set empty. This bug probably also applied to
6024 41. The strip_trailing_dot option was not being applied to the address given
6025 with the -f command-line option.
6027 42. The code for reading a message's header from the spool was incrementing
6028 $received_count, but never initializing it. This meant that the value was
6029 incorrect (doubled) while delivering a message in the same process in which
6030 it was received. In the most common configuration of Exim, this never
6031 happens - a fresh exec is done - but it can happen when
6032 deliver_drop_privilege is set.
6034 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6035 it now includes up to 150 characters of the unexpected data in the log
6038 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6039 and building data strings. The size of both of these buffers was 10 000
6040 bytes - far larger than anybody would *ever* want, thought I. Needless to
6041 say, somebody hit the limit. I have increased the maximum line length to
6042 20 000 and the maximum data length of concatenated lines to 100 000. I have
6043 also fixed two bugs, because there was no checking on these buffers. Tsk,
6044 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6045 buffer is too small.
6047 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6048 lsearch lookups. Now it does.
6050 46. When parsing a route_list item in a manualroute router, a fixed-length
6051 buffer was used for the list of hosts. I made this 1024 bytes long,
6052 thinking that nobody would ever have a list of hosts that long. Wrong.
6053 Somebody had a whole pile of complicated expansion conditions, and the
6054 string was silently truncated, leading to an expansion error. It turns out
6055 that it is easier to change to an unlimited length (owing to other changes
6056 that have happened since this code was originally written) than to build
6057 structure for giving a limitation error. The length of the item that
6058 expands into the list of hosts is now unlimited.
6060 47. The lsearch lookup could not handle data where the length of text line was
6061 more than 4095 characters. Such lines were truncated, leading to shortened
6062 data being returned. It should now handle lines of any length.
6064 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6065 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6068 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6069 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6070 no longer happens in comment lines. A list of replacements is now placed
6071 at the head of all of the source files, except those whose only change is
6072 to replace PERL_COMMAND in the very first #! line.
6074 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6075 messages on the queue, with a bottom-up merge sort, using code contributed
6076 by Michael Haardt. This should make operations like -bp somewhat faster on
6077 large queues. It won't affect queue runners, except when queue_run_in_order
6080 51. Installed eximstats 1.31 in the distribution.
6082 52. Added support for SRV lookups to the dnslookup router.
6084 53. If an ACL referred to $message_body or $message_body_end, the value was not
6085 reset for any messages that followed in the same SMTP session.
6087 54. The store-handling optimization for building very long strings was not
6088 differentiating between the different store pools. I don't think this
6089 actually made any difference in practice, but I've tidied it.
6091 55. While running the routers to verify a sender address, $sender_address
6092 was still set to the sender address. This is wrong, because when routing to
6093 send a bounce to the sender, it would be empty. Therefore, I have changed
6094 it so that, while verifying a sender address, $sender_address is set to <>.
6095 (There is no change to what happens when verifying a recipient address.)
6097 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6098 target A or AAAA records (if not already returned) without resetting the
6099 qualify_single or search_parents options of the DNS resolver. These are
6100 inappropriate in this case because the targets of MX and SRV records must
6101 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6102 target that, when qualified, matched something in the local domain. These
6103 two options are now turned off when doing these lookups.
6105 57. It seems that at least some releases of Reiserfs (which does not have the
6106 concept of a fixed number of inodes) returns zero and not -1 for the
6107 number of available inodes. This interacted badly with check_spool_inodes,
6108 which assumed that -1 was the "no such thing" setting. What I have done is
6109 to check that the total number of inodes is greater than zero before doing
6110 the test of how many are available.
6112 58. When a "warn" ACL statement has a log_message modifier, the message is
6113 remembered, and not repeated. This is to avoid a lot of repetition when a
6114 message has many recipients that cause the same warning to be written.
6115 However, Exim was preserving the list of already written lines for an
6116 entire SMTP session, which doesn't seem right. The memory is now reset if a
6117 new message is started.
6119 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6120 debugging output unless log_rewrite was also set.
6122 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6123 of (int)(handle) when we know that handle contains (void *)(-1).
6125 61. The Exim daemon panic-logs an error return when it closes the incoming
6126 connection. However "connection reset by peer" seems to be common, and
6127 isn't really an error worthy of noting specially, so that particular error
6130 62. When Exim is trying to find all the local interfaces, it used to panic and
6131 die if the ioctl to get the interface flags failed. However, it seems that
6132 on at least one OS (Solaris 9) it is possible to have an interface that is
6133 included in the list of interfaces, but for which you get a failure error
6134 for this call. This happens when the interface is not "plumbed" into a
6135 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6136 failure of the "get flags" call assumes that the interface is down.
6138 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6139 makes life easier for people who are doing arithmetic on fields extracted
6140 from dates, where you often get leading zeros that should not be
6141 interpreted as octal.
6143 64. Added qualify_domain to the redirect router, to override the global
6146 65. If a pathologically long header line contained very many addresses (the
6147 report of this problem mentioned 10 000) and each of them was rewritten,
6148 Exim could use up a very large amount of memory. (It kept on making new
6149 copies of the header line as it rewrote, and never released the old ones.)
6150 At the expense of a bit more processing, the header rewriting function has
6151 been changed so that it no longer eats memory in this way.
6153 66. The generation of the Received: header has been moved from the time that a
6154 message starts to be received, to the time that it finishes. The timestamp
6155 in the Received: header should now be very close to that of the <= log
6156 line. There are two side-effects of this change:
6158 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6159 logged header lines no longer include the local Received: line, because
6160 it has not yet been created. The same applies to a copy of the message
6161 that is returned to a non-SMTP sender when a message is rejected.
6163 (b) When a filter file is tested using -bf, no additional Received: header
6164 is added to the test message. After some thought, I decided that this
6167 This change does not affect the value of $received_for. It is still set
6168 after address rewriting, but before local_scan() is called.
6170 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6172 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6173 gave an unhelpful panic error message, and a defer error. I have managed to
6174 change this behaviour so that it now rejects any supplied certificate,
6175 which seems right, as the list of acceptable certificates is empty.
6177 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6178 gave an unhelpful defer error. I have not managed to make this reject any
6179 supplied certificates, but the error message it gives is "no certificate
6180 supplied", which is not helpful.
6182 70. exigrep's output now also includes lines that are not associated with any
6183 message, but which match the given pattern. Implemented by a patch from
6184 Martin Sluka, which also tidied up the Perl a bit.
6186 71. Recipient callout verification, like sender verification, was using <> in
6187 the MAIL FROM command. This isn't really the right thing, since the actual
6188 sender may affect whether the remote host accepts the recipient or not. I
6189 have changed it to use the actual sender in the callout; this means that
6190 the cache record is now keyed on a recipient/sender pair, not just the
6191 recipient address. There doesn't seem to be a real danger of callout loops,
6192 since a callout by the remote host to check the sender would use <>.
6193 [SEE ABOVE: changed after hitting problems.]
6195 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6196 temporary errors. However, in the case of such a code being given after
6197 the end of a data transmission (i.e. after ".") Exim was failing to write
6198 a retry record for the message. (Yes, there was some broken host that was
6199 actually sending 8xx at this point.)
6201 73. An unknown lookup type in a host list could cause Exim to panic-die when
6202 the list was checked. (An example that provoked this was putting <; in the
6203 middle of a list instead of at the start.) If this happened during a DATA
6204 ACL check, a -D file could be left lying around. This kind of configuration
6205 error no longer causes Exim to die; instead it causes a defer error. The
6206 incident is still logged to the main and panic logs.
6208 74. Buglet left over from Exim 3 conversion. The message "too many messages
6209 in one connection" was written to the rejectlog but not the mainlog, except
6210 when address rewriting (yes!) was being logged.
6212 75. Added write_rejectlog option.
6214 76. When a system filter was run not as root (that is, when system_filter_user
6215 was set), the values of the $n variables were not being returned to the
6216 main process; thus, they were not subsequently available in the $sn
6219 77. Added +return_path_on_delivery log selector.
6221 78. A connection timeout was being treated differently from recipients deferred
6222 when testing hosts_max_try with a message that was older than the host's
6223 retry timeout. (The host should not be counted, thus allowing all hosts to
6224 be tried at least once before bouncing.) This may have been the cause of an
6225 occasionally reported bug whereby a message would remain on the queue
6226 longer than the retry timeout, but would be bounced if a delivery was
6227 forced. I say "may" because I never totally pinned down the problem;
6228 setting up timeout/retry tests is difficult. See also the next item.
6230 79. The ultimate address timeout was not being applied to errors that involved
6231 a combination of host plus message (for example, a timeout on a MAIL
6232 command). When an address resolved to a number of possible hosts, and they
6233 were not all tried for each delivery (e.g. because of hosts_max_try), a
6234 message could remain on the queue longer than the retry timeout.
6236 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6239 81. Fixed an obscure SMTP outgoing bug which required at least the following
6240 conditions: (a) there was another message waiting for the same server;
6241 (b) the server returned 5xx to all RCPT commands in the first message so
6242 that the message was not completed; (c) the server dropped the connection
6243 or gave a negative response to the RSET that Exim sends to abort the
6244 transaction. The observed case was a dropped connection after DATA that had
6245 been sent in pipelining mode. That is, the server had advertised PIPELINING
6246 but was not implementing it correctly. The effect of the bug was incorrect
6247 behaviour, such as trying another host, and this could lead to a crash.
6253 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6254 and daemon.c were passed as pointers to ints; they should have been
6255 pointers to socklen_t variables (which are typically unsigned ints).
6257 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6260 3. Fixed a really odd bug that affected only the testing scheme; patching a
6261 certain fixed string in the binary changed the value of another string that
6262 happened to be identical to the end of the original first string.
6264 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6265 name", it returns that address as the IP address. On some operating
6266 systems (e.g. Solaris), it also passes back the IP address string as the
6267 "host name". However, on others (e.g. Linux), it passes back an empty
6268 string. Exim wasn't checking for this, and was changing the host name to an
6269 empty string, assuming it had been canonicalized.
6271 5. Although rare, it is permitted to have more than one PTR record for a given
6272 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6273 all the names associated with an address, because they do in Solaris.
6274 However, it seems that they do not in Linux for data that comes from the
6275 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6276 I found this out when I moved to a new Linux workstation and tried to run
6277 the Exim test suite.
6279 To get round this problem I have changed the code so that it now does its
6280 own call to the DNS to look up PTR records when searching for a host name.
6281 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6282 addresses that are only in /etc/hosts are still found.
6284 This behaviour is, however, controlled by an option called host_lookup_
6285 order, which defaults to "bydns:byaddr". If people want to use the other
6286 order, or indeed, just use one or the other means of lookup, they can
6287 specify it in this variable.
6289 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6290 some operating systems, this comes back from gethostbyaddr() as an empty
6291 string, and this is what Exim used to test for. However, it seems that in
6292 other systems, "." is yielded. Exim now tests for this case too.
6294 7. The values of check_spool_space and check_log_space are now held internally
6295 as a number of kilobytes instead of an absolute number of bytes. If a
6296 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6297 kilobyte. This means that much larger values can be stored.
6299 8. Exim monitor: an attempt to get the action menu when not actually pointing
6300 at a message produces an empty menu entitled "No message selected". This
6301 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6302 no entries in it ("Shell widget menu has zero width and/or height"). So I
6303 have added a single, blank menu entry in this case.
6305 9. Added ${quote_local_part.
6307 10. MIME decoding is now applied to the contents of Subject: header lines when
6310 11. Now that a reference to $sender_host_address automatically causes a reverse
6311 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6312 host lookup before query-style lookups in lists that might use this
6313 variable. This has therefore been abolished, and the "net-" prefix is no
6314 longer necessary for query-style lookups.
6316 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6317 have been a typo for LFLAGS, so it has been changed.
6319 13. The install script calls Exim with "-C /dev/null" in order to find the
6320 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6321 to be output. However, since Exim outputs its version number before the
6322 error, it didn't break the script. It just looked ugly. I fixed this by
6323 always allowing "-C /dev/null" if the caller is root.
6325 14. Ignore overlarge ACL variable number when reading spool file - insurance
6326 against a later release with more variables having written the file.
6328 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6329 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6330 this, as well as the form without the "IPv6" on the front.
6332 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6333 OS/Makefile-Darwin file.
6335 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6336 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6338 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6339 variable. However, some OS use ints for inodes. I've added cast to long int
6340 to get rid of the compiler warning.
6342 19. I had forgotten to lock out "/../" in configuration file names when
6343 ALT_CONFIG_PREFIX was set.
6345 20. Routers used for verification do not need to specify transports. However,
6346 if such a router generated a host list, and callout was configured, Exim
6347 crashed, because it could not find a port number from the (non-existent)
6348 transport. It now assumes port 25 in this circumstance.
6350 21. Added the -t option to exigrep.
6352 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6353 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6354 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6355 them is set, LOOKUP_LSEARCH is forced.
6357 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6359 24. Added sender and host information to the "rejected by local_scan()" log
6360 line; previously there was no indication of these.
6362 25. Added .include_if_exists.
6364 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6365 Linux. It turns out that not all file systems support this. Apparently some
6366 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6367 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6368 not supported on the file descriptor, is now ignored when Exim is trying to
6369 sync a directory. This applies only to Linux.
6371 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6373 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6374 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6377 29. Added support for maildirsize files from supplied patch (modified a bit).
6379 30. The use of :fail: followed by an empty string could lead Exim to respond to
6380 sender verification failures with (e.g.):
6382 550 Verification failed for <xxx>
6383 550 Sender verify failed
6385 where the first response line was missing the '-' that indicates it is not
6386 the final line of the response.
6388 31. The loop for finding the name of the user that called Exim had a hardwired
6389 limit of 10; it now uses the value of finduser_retries, which is used for
6390 all other user lookups.
6392 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6395 33. Exim was neglecting to zero errno before one call of strtol() when
6396 expanding a string and expecting an integer value. On some systems this
6397 resulted in spurious "integer overflow" errors. Also, it was casting the
6398 result into an int without checking.
6400 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6401 did not work. The code looks as if it has *never* worked, though it appears
6402 to have been documented since at least release 1.62. I have made it work.
6404 35. The "timeout_DNS" error in retry rules, also documented since at least
6405 1.62, also never worked. As it isn't clear exactly what this means, and
6406 clearly it isn't a major issue, I have abolished the feature by treating it
6407 as "timeout", and writing a warning to the main and panic logs.
6409 36. The display of retry rules for -brt wasn't always showing the error code
6412 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6413 timeout_connect_A, timeout_connect_MX.
6415 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6416 to the empty sender.
6418 39. The daemon was not analysing the content of -oX till after it had closed
6419 stderr and disconnected from the controlling terminal. This meant that any
6420 syntax errors were only noted on the panic log, and the return code from
6421 the command was 0. By re-arranging the code a little, I've made the
6422 decoding happen first, so such errors now appear on stderr, and the return
6423 code is 1. However, the actual setting up of the sockets still happens in
6424 the disconnected process, so errors there are still only recorded on the
6427 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6428 connections (as happens on some IP stacks) was logged at start up time as
6429 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6430 it from "IPv6 and IPv4", which means that two separate sockets are being
6433 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6434 says whether AF_INET or AF_INET6 was passed as an argument.
6436 42. Exiwhat output was messed up when time zones were included in log
6439 43. Exiwhat now gives more information about the daemon's listening ports,
6440 and whether -tls-on-connect was used.
6442 44. The "port" option of the smtp transport is now expanded.
6444 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6445 silently ignored. Now an error message is written to the main and panic
6448 46. There's a new ACL modifier called "logwrite" which writes to a log file
6449 as soon as it is encountered.
6451 47. Added $local_user_uid and $local_user_gid at routing time.
6453 48. Exim crashed when trying to verify a sender address that was being
6456 49. Exim was recognizing only a space character after ".include". It now also
6457 recognizes a tab character.
6459 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6460 extracting the relevant information from the specification. The man page no
6461 longer contains scrambled data for the -d option, and I've added a section
6462 at the front about calling Exim under different names.
6464 51. Added "extra_headers" argument to the "mail" command in filter files.
6466 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6469 53. Installed eximstats 1.29.
6471 54. Added transport_filter_timeout as a generic transport option.
6473 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6474 Cc: header lines. This was required by RFC 822, but it not required by RFC
6477 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6478 incoming messages that did not have them. Now it does so only if the
6479 message originates locally, that is, if there is no associated remote host
6480 address. When Resent- header lines are present, this applies to the Resent-
6481 lines rather than the non-Resent- lines.
6483 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6484 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6486 58. Messages for configuration errors now include the name of the main
6487 configuration file - useful now that there may be more than one file in a
6488 list (.included file names were always shown).
6490 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6491 for those rare installations that do not start the daemon as root or run it
6492 setuid root. I've cut out the call to initgroups() if the daemon is not
6495 60. The Exim user and group can now be bound into the binary as text strings
6496 that are looked up at the start of Exim's processing.
6498 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6500 62. Added $mailstore_basename variable.
6502 63. Installed patch to sieve.c from Michael Haardt.
6504 64. When Exim failed to open the panic log after failing to open the main log,
6505 the original message it was trying to log was written to stderr and debug
6506 output, but if they were not available (the usual case in production), it
6507 was lost. Now it is written to syslog before the two lines that record the
6508 failures to open the logs.
6510 65. Users' Exim filters run in subprocesses under the user's uid. It is
6511 possible for a "deliver" command or an alias in a "personal" command to
6512 provoke an address rewrite. If logging of address rewriting is configured,
6513 this fails because the process is not running as root or exim. There may be
6514 a better way of dealing with this, but for the moment (because 4.30 needs
6515 to be released), I have disabled address rewrite logging when running a
6516 filter in a non-root, non-exim process.
6522 1. The buildconfig auxiliary program wasn't quoting the value set for
6523 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6524 not defined. This bug was masked in 4.22 by the effect that was fixed in
6527 2. Some messages that were rejected after a message id was allocated were
6528 shown as "incomplete" by exigrep. It no longer does this for messages that
6529 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6531 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6532 have allow_domain_literals set, the ID did not get logged in the <= line.
6533 Domain literals are now always recognized in Message-ID: header lines.
6535 4. The first argument for a ${extract expansion item is the key name or field
6536 number. Leading and trailing spaces in this item were not being ignored,
6537 causing some misleading effects.
6539 5. When deliver_drop_privilege was set, single queue runner processes started
6540 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6541 same command in the process it spins off) were not dropping privilege.
6543 6. When the daemon running as "exim" started a queue runner, it always
6544 re-executed Exim in the spun-off process. This is a waste of effort when
6545 deliver_drop_privilege is set. The new process now just calls the
6546 queue-runner function directly.
6552 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6555 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6556 ignored. Though the use of -oP was forcing the writing of a pid file, it
6557 was always written to the default place.
6559 3. If the message "no IP address found for host xxxx" is generated during
6560 incoming verification, it is now followed by identification of the incoming
6561 connection (so you can more easily find what provoked it).
6563 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6565 5. Added some features to "harden" Exim a bit more against certain attacks:
6567 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6568 be put in Local/Makefile. This is like the never_users runtime option,
6569 but it cannot be overridden. The default setting is "root".
6571 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6572 prefix string with which any file named in a -C command line option
6575 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6576 is retained for -C and -D only if the caller of Exim is root. Without
6577 it, the exim user may also use -C and -D and retain privilege.
6579 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6580 command line option is disabled.
6582 6. Macro names set by the -D option must start with an upper case letter, just
6583 like macro names defined in the configuration file.
6585 7. Added "dereference=" facility to LDAP.
6587 8. Two instances of the typo "uknown" in the source files are fixed.
6589 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6590 the Configure-Makefile script screwed up while processing it.
6592 10. Incorporated PCRE 4.4.
6594 11. The SMTP synchronization check was not operating right at the start of an
6595 SMTP session. For example, it could not catch a HELO sent before the client
6596 waited for the greeting. There is now a check for outstanding input at the
6597 point when the greeting is written. Because of the duplex, asynchronous
6598 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6599 way, but not yet received, when the check is performed.
6601 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6602 on TCP/IP sockets, because this apparently causes some broken clients to
6605 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6606 unchanged) from the Cygwin maintainer.
6608 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6609 was supported instead of testing for mbx. Effectively a typo.
6611 15. The spa authenticator server code was not checking that the input it
6612 received was valid base64.
6614 16. The debug output line for the "set" modifier in ACLs was not showing the
6615 name of the variable that was being set.
6617 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6619 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6620 lookup on its own. Something else was needed to trigger the lookup. For
6621 example, a match in host_lookup or the need for a host name in a host list.
6622 Now, if $sender_host_name is referenced and the host name has not yet been
6623 looked up, a lookup is performed. If the lookup fails, the variable remains
6624 empty, and $host_lookup_failed is set to "1".
6626 19. Added "eqi" as a case-independent comparison operator.
6628 20. The saslauthd authentication condition could segfault if neither service
6629 nor realm was specified.
6631 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6632 error message that was logged was misleading, and incoming SMTP
6633 connections were dropped. The message is now more accurate, and temporary
6634 errors are given to SMTP connections.
6636 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6637 (except RSET) with a 421 error, until QUIT is received. However, it was
6638 failing to send a response to QUIT.
6640 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6641 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6642 ACL is now run after the helo_try_verify_hosts code.
6644 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6645 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6646 case-independent, so other case variants are also recognized). Apparently
6647 some systems use these upper case variants.
6649 25. If more than two messages were waiting for the same host, and a transport
6650 filter was specified for the transport, Exim sent two messages over the
6651 same TCP/IP connection, and then failed with "socket operation on non-
6652 socket" when it tried to send the third.
6654 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6656 27. The extern definition of crypt16() in expand.c was not being excluded when
6657 the OS had its own crypt16() function.
6659 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6660 as a preferred synonym for return_size_limit, both as an option and as an
6663 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6665 30. Changed the default configuration ACL to relax the local part checking rule
6666 for addresses that are not in any local domains. For these addresses,
6667 slashes and pipe symbols are allowed within local parts, but the sequence
6668 /../ is explicitly forbidden.
6670 31. SPA server authentication was not clearing the challenge buffer before
6673 32. log_message in a "warn" ACL statement was writing to the reject log as
6674 well as to the main log, which contradicts the documentation and doesn't
6675 seem right (because no rejection is happening). So I have stopped it.
6677 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6678 However, I am unable to do any testing of this.
6680 34. Fixed an infelicity in the appendfile transport. When checking directories
6681 for a mailbox, to see if any needed to be created, it was accidentally
6682 using path names with one or more superfluous leading slashes; tracing
6683 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6685 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6686 contents are added to the log line that is written for every discarded
6687 recipient. (Previously a log_message setting was ignored.)
6689 36. The ${quote: operator now quotes the string if it is empty.
6691 37. The install script runs exim in order to find its version number. If for
6692 some reason other than non-existence or emptiness, which it checks, it
6693 could not run './exim', it was installing it with an empty version number,
6694 i.e. as "exim-". This error state is now caught, and the installation is
6697 38. An argument was missing from the function that creates an error message
6698 when Exim fails to connect to the socket for saslauthd authentication.
6699 This could cause Exim to crash, or give a corrupted message.
6701 39. Added isip, isip4, and isip6 to ${if conditions.
6703 40. The ACL variables $acl_xx are now saved with the message, and can be
6704 accessed later in routers, transports, and filters.
6706 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6707 strings in the file are not string-expanded.
6709 42. If a MAIL command specified a SIZE value that was too large to fit into an
6710 int variable, the check against message_size_limit failed. Such values are
6711 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6712 day this will have to be increased, but I don't think I want to be around
6713 when emails are that large.
6720 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6721 iconv() is not standard in FreeBSD.
6723 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6724 IPv6 enabled. The observed symptom was a segmentation fault on return from
6725 the function os_common_find_running_interfaces() in src/os.c.
6727 3. In the check_special_case() function in daemon.c I had used "errno" as an
6728 argument name, which causes warnings on some systems. This was basically a
6729 typo, since it was named "eno" in the comments!
6731 4. The code that waits for the clock to tick (at a resolution of some fraction
6732 of a second) so as to ensure message-id uniqueness was always waiting for
6733 at least one whole tick, when it could have waited for less. [This is
6734 almost certainly not relevant at current processor speeds, where it is
6735 unlikely to ever wait at all. But we try to future-proof.]
6737 5. The function that sleeps for a time interval that includes fractions of a
6738 second contained a race. It did not block SIGALRM between setting the
6739 timer, and suspending (a couple of lines later). If the interval was short
6740 and the sigsuspend() was delayed until after it had expired, the suspension
6741 never ended. On busy systems this could lead to processes getting stuck for
6744 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6745 process, before it forks any delivery processes. The open lookup caching
6746 mechanism meant that the open file or database connection was passed into
6747 the delivery process. The problem was that delivery processes always tidy
6748 up cached lookup data. This could cause a problem for the next delivery
6749 process started by the queue runner, because the external queue runner
6750 process does not know about the closure. So the next delivery process
6751 still has data in the lookup cache. In the case of a file lookup, there was
6752 no problem because closing a file descriptor in a subprocess doesn't affect
6753 the parent. However, if the lookup was caching a connection to a database,
6754 the connection was closed, and the second delivery process was likely to
6755 see errors such as "PGSQL: query failed: server closed the connection
6756 unexpectedly". The problem has been fixed by closing all cached lookups
6757 in a queue runner before running a delivery process.
6759 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6760 seem to have the "const" qualifier which it has on other OS. I've
6763 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6764 *of the same type* (client or server) with the same public name that an
6765 error should be diagnosed.
6767 9. When Exim looked up a host name for an IP address, but failed to find the
6768 original IP address when looking up the host name (a safety check), it
6769 output the message "<ip address> does not match any IP for NULL", which was
6770 confusing, to say the least. The bug was that the host name should have
6771 appeared instead of "NULL".
6773 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6774 user that is built into the binary, and the -C or -D options is used, root
6775 privilege is dropped before the configuration file is read. In addition,
6776 logging is switched to stderr instead of the normal log files. If the
6777 configuration then re-defines the Exim user, the unprivileged environment
6778 is probably not what is expected, so Exim logs a panic warning message (but
6781 However, if deliver_drop_privilege is set, the unprivileged state may well
6782 be exactly what is intended, so the warning has been cut out in that case,
6783 and Exim is allowed to try to write to its normal log files.
6789 1. smtp_return_error_details was not giving details for temporary sender
6790 or receiver verification errors.
6792 2. Diagnose a configuration error if two authenticators have the same public
6795 3. Exim used not to create the message log file for a message until the first
6796 delivery attempt. This could be confusing when incoming messages were held
6797 for policy or load reasons. The message log file is now created at the time
6798 the message is received, and an initial "Received" line is written to it.
6800 4. The automatically generated man page for command line options had a minor
6801 bug that caused no ill effects; however, a more serious problem was that
6802 the procedure for building the man page automatically didn't always
6803 operate. Consequently, release 4.20 contains an out-of-date version. This
6804 shouldn't happen again.
6806 5. When building Exim with embedded Perl support, the script that builds the
6807 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6808 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6810 6. The freeze_tell option was not being used for messages that were frozen on
6811 arrival, either by an ACL or by local_scan().
6813 7. Added the smtp_incomplete_transaction log selector.
6815 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6816 was accepting AUTH without a new EHLO.
6818 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6819 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6822 10. Logging of TCP/IP connections (when configured) now happens in the main
6823 daemon process instead of the child process, so that the TCP/IP connection
6824 count is more accurate (but it can never be perfect).
6826 11. The use of "drop" in a nested ACL was not being handled correctly in the
6827 outer ACL. Now, if condition failure induced by the nested "drop" causes
6828 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6829 or "require"), the connection is dropped.
6831 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6832 that yield "discard" can now be used with an "accept" or a "discard" verb,
6833 but an error is generated for any others (because I can't see a useful way
6834 to define what should happen).
6836 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6837 are now processed in the same way as lines in the Exim configuration file.
6838 In particular, continuation lines are supported.
6840 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6842 15. Added -ti meaning -t -i.
6844 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6845 domains, and warn by logging if others are found.
6847 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6848 returned from GIFCONF when Exim is trying to find the list of interfaces on
6849 a host. The code in os.c has been modified to copy each ifreq to an aligned
6850 structure in all cases.
6852 Also, in some cases, the returned ifreq's were being copied to a 'struct
6853 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6854 means the last couple of bytes of an IPv6 address could be chopped if the
6855 ifreq contained only a normal sockaddr (14 bytes storage).
6857 18. Named domain lists were not supported in the hosts_treat_as_local option.
6858 An entry such as +xxxx was not recognized, and was treated as a literal
6861 19. Ensure that header lines added by a DATA ACL are included in the reject log
6862 if the ACL subsequently rejects the message.
6864 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6865 MD5 (which is deprecated).
6867 21. When testing a filter file using -bf, Exim was writing a message when it
6868 took the sender from a "From " line in the message, but it was not doing so
6869 when it took $return_path from a Return-Path: header line. It now does.
6871 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6872 with a valid header line field name (a series of printing characters
6873 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6875 23. Changed "disc" in the source to "disk" to conform to the documentation and
6876 the book and for uniformity.
6878 24. Ignore Sendmail's -Ooption=value command line item.
6880 25. When execve() failed while trying to run a command in a pipe transport,
6881 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6882 could be confused with a return value of 69 from the command itself. This
6883 has been changed to 127, the value the shell returns if it is asked to run
6884 a non-existent command. The wording for the related log line suggests a
6885 non-existent command as the problem.
6887 26. If received_header_text expands to an empty string, do not add a Received:
6888 header line to the message. (Well, it adds a token one on the spool, but
6889 marks it "old" so that it doesn't get used or transmitted.)
6891 27. Installed eximstats 1.28 (addition of -nt option).
6893 28. There was no check for failure on the call to getsockname() in the daemon
6894 code. This can fail if there is a shortage of resources on the system, with
6895 ENOMEM, for example. A temporary error is now given on failure.
6897 29. Contrary to the C standard, it seems that in some environments, the
6898 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6899 program. Exim now does this explicitly; it affects the formatting of
6900 timestamps using strftime().
6902 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6903 complaints. I've now initialized all the variables, to avoid this.
6905 32. Header lines added by a system filter were not being "seen" during
6906 transport-time rewrites.
6908 33. The info_callback() function passed to OpenSSL is set up with type void
6909 (*)(SSL *, int, int), as described somewhere. However, when calling the
6910 function (actually a macro) that sets it up, the type void(*)() is
6911 expected. I've put in a cast to prevent warnings from picky compilers.
6913 34. If a DNS black list lookup found a CNAME record, but there were no A
6914 records associated with the domain it pointed at, Exim crashed.
6916 35. If a DNS black list lookup returned more than one A record, Exim ignored
6917 all but the first. It now scans all returned addresses if a particular IP
6918 value is being sought. In this situation, the contents of the
6919 $dnslist_value variable are a list of all the addresses, separated by a
6922 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6923 to accept a host name and all its aliases if the forward lookup for any of
6924 them yielded the IP address of the incoming connection. Now it accepts only
6925 those names whose forward lookup yields the correct IP address. Any other
6926 names are discarded. This closes a loophole whereby a rogue DNS
6927 administrator could create reverse DNS records to break through a
6928 wildcarded host restriction in an ACL.
6930 37. If a user filter or a system filter that ran in a subprocess used any of
6931 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6932 the wrong values were passed to the pipe command ($thisaddress had the
6933 value of $0, $0 had the value of $1, etc). This bug was introduced by
6934 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6936 38. Improved the line breaking for long SMTP error messages from ACLs.
6937 Previously, if there was no break point between 40 and 75 characters, Exim
6938 left the rest of the message alone. Two changes have been made: (a) I've
6939 reduced the minimum length to 35 characters; (b) if it can't find a break
6940 point between 35 and 75 characters, it looks ahead and uses the first one
6941 that it finds. This may give the occasional overlong line, but at least the
6942 remaining text gets split now.
6944 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6945 file descriptors might be low, and that setting 1000 would always raise it.
6946 It turns out that in some environments, the limit is already over 1000 and
6947 that lowering it causes trouble. So now Exim takes care not to decrease it.
6949 40. When delivering a message, the value of $return_path is set to $sender_
6950 address at the start of routing (routers may change the value). By an
6951 oversight, this default was not being set up when an address was tested by
6952 -bt or -bv, which affected the outcome if any router or filter referred to
6955 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6956 log only when "message" or "log_message" are set. However, if one of the
6957 conditions was an address verification, or a call to a nested ACL, the
6958 messages generated by the underlying test were being passed through. This
6959 no longer happens. The underlying message is available in $acl_verify_
6960 message for both "message" and "log_message" expansions, so it can be
6961 passed through if needed.
6963 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6964 new expansion $bh_ to give the encoded byte string without charset
6965 translation. Translation happens only if iconv() is available; HAVE_ICONV
6966 indicates this at build time. HEADERS_CHARSET gives the charset to
6967 translate to; headers_charset can change it in the configuration, and
6968 "headers charset" can change it in an individual filter file.
6970 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6971 that creates RFC 2047 encoded "words" labels them as that charset instead
6972 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6973 expansion operator; (ii) when Exim creates a From: line for a local
6974 message; (iii) when a header line is rewritten to include a "phrase" part.
6976 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6977 buggy, causing it to skip the first lines of messages whose message ID
6978 ended in 'D'. This would not have bitten before Exim release 4.14, because
6979 message IDs were unlikely to end in 'D' before then. The effect was to have
6980 incorrect size information for certain domains.
6982 45. #include "config.h" was missing at the start of the crypt16.c module. This
6983 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6986 46. If there was a timeout during a "random" callout check, Exim treated it as
6987 a failure of the random address, and carried on sending RSET and the real
6988 address. If the delay was just some slowness somewhere, the response to the
6989 original RCPT would be taken as a response to RSET and so on, causing
6990 mayhem of various kinds.
6992 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6993 when I implemented it. It didn't allow for the fact that some option values
6994 may legitimately be negative (e.g. size_addition), and it didn't even do
6995 the right test for positive values.
6997 48. Domain names in DNS records are case-independent. Exim always looks them up
6998 in lower case. Some resolvers return domain names in exactly the case they
6999 appear in the zone file, that is, they may contain uppercase letters. Not
7000 all resolvers do this - some return always lower case. Exim was treating a
7001 change of case by a resolver as a change of domain, similar to a widening
7002 of a domain abbreviation. This triggered its re-routing code and so it was
7003 trying to route what was effectively the same domain again. This normally
7004 caused routing to fail (because the router wouldn't handle the domain
7005 twice). Now Exim checks for this case specially, and just changes the
7006 casing of the domain that it ultimately uses when it transmits the message
7009 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7012 50. If a filter generated a file delivery with a non-absolute name (possible if
7013 no home directory exists for the router), the forbid_file option was not
7016 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7017 the existing equality matching.
7019 52. Exim was using ints instead of ino_t variables in some places where it was
7020 dealing with inode numbers.
7022 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7023 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7024 called TMPDIR, and if it finds it is different, it changes its value.
7026 54. The smtp_printf() function is now made available to local_scan() so
7027 additional output lines can be written before returning. There is also an
7028 smtp_fflush() function to enable the detection of a dropped connection.
7029 The variables smtp_input and smtp_batched_input are exported to
7032 55. Changed the default runtime configuration: the message "Unknown user"
7033 has been removed from the ACL, and instead placed on the localuser router,
7034 using the cannot_route_message feature. This means that any verification
7035 failures that generate their own messages won't get overridden. Similarly,
7036 the "Unrouteable address" message that was in the ACL for unverifiable
7037 relay addresses has also been removed.
7039 56. Added hosts_avoid_esmtp to the smtp transport.
7041 57. The exicyclog script was not checking for the esoteric option
7042 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7043 will work only if exicyclog is run under the appropriate euid.
7045 58. Following a discussion on the list, the rules by which Exim recognises line
7046 endings on incoming messages have been changed. The -dropcr and drop_cr
7047 options are now no-ops, retained only for backwards compatibility. The
7048 following line terminators are recognized: LF CRLF CR. However, special
7049 processing applies to CR:
7051 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7052 nor a local message in the state where . is a terminator.
7054 (ii) If a bare CR is encountered in a header line, an extra space is added
7055 after the line terminator so as not to end the header. The reasoning
7056 behind this is that bare CRs in header lines are most likely either
7057 to be mistakes, or people trying to play silly games.
7059 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7060 was being incorrectly given as 18 bytes larger than it should have been.
7061 This is a VOB (very old bug).
7063 60. This may never have affected anything current, but just in case it has:
7064 When the local host is found other than at the start of a list of hosts,
7065 the local host, those with the same MX, and any that follow, are discarded.
7066 When the list in question was part of a longer list of hosts, the following
7067 hosts (not currently being processed) were also being discarded. This no
7068 longer happens. I'm not sure if this situation could ever has previously
7071 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7074 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7075 header. This is something that is recommended in a new Internet Draft, and
7076 is something that is documented as being done by Sendmail. There are two
7077 possible values. For messages generated by the autoreply transport, Exim
7080 Auto-Submitted: auto-replied
7082 whereas for all other generated messages (e.g. bounces) it adds
7084 Auto-Submitted: auto-generated
7086 63. The "personal" condition in filters now includes a test for the
7087 Auto-Submitted: header. If it contains the string "auto-" the message it
7088 not considered personal.
7090 64. Added rcpt_include_affixes as a generic transport option.
7092 65. Added queue_only_override (default true).
7094 66. Added the syslog_duplication option.
7096 67. If what should have been the first header line of a message consisted of
7097 a space followed by a colon, Exim was mis-interpreting it as a header line.
7098 It isn't of course - it is syntactically invalid and should therefore be
7099 treated as the start of the message body. The misbehaviour could have
7100 caused a number of strange effects, including loss of data in subsequent
7101 header lines, and spool format errors.
7103 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7104 client host had authenticated. This control can now be exercised by an ACL
7105 for more flexibility.
7107 69. By default, callouts do not happen when testing with -bh. There is now a
7108 variant, -bhc, which does actually run the callout code, including
7109 consulting and updating the callout cache.
7111 70. Added support for saslauthd authentication, courtesy of Alexander
7114 71. If statvfs() failed on the spool or log directories while checking their
7115 size for availability, Exim confusingly gave the error "space shortage".
7116 Furthermore, in debugging mode it crashed with a floating point exception.
7117 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7118 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7119 serious problem, Exim now writes to the main and panic logs when this
7120 happens, with details of the failure. It then refuses to accept the
7121 incoming message, giving the message "spool directory problem" or "log
7122 directory problem" with a 421 code for SMTP messages.
7124 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7125 0, 1, and 2 exist, because some OS complain for execs without them (see
7126 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7127 descriptors. However, the code omitted to check that the open succeeded,
7128 causing mysterious errors if for some reason the permissions on /dev/null
7129 got screwed. Now Exim writes a message to the main and panic logs, and
7130 bombs out if it can't open /dev/null.
7132 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7133 interact so that it is all more flexible. It is supposed to remain
7134 backwards compatible. Also added extra_local_interfaces.
7136 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7137 to bomb out with an assertion failure - to the client this appears as a
7138 connection drop. This problem occurs in the part of the code that was taken
7139 from the Samba project. Fortunately, the assertion is in a very simple
7140 function, so I have fixed this by reproducing the function inline in the
7141 one place where it is called, and arranging for authentication to fail
7142 instead of killing the process with assert().
7144 75. The SPA client code was not working when the server requested OEM rather
7145 than Unicode encoding.
7147 76. Added code to make require_files with a specific uid setting more usable in
7148 the case where statting the file as root fails - usually a non-root-mounted
7149 NFS file system. When this happens and the failure is EACCES, Exim now
7150 forks a subprocess and does the per-uid checking as the relevant uid.
7152 77. Added process_log_path.
7154 78. If log_file_path was not explicitly set, a setting of check_log_space or
7155 check_log_inodes was ignored.
7157 79. If a space check for the spool or log partitions fails, the incident is now
7158 logged. Of course, in the latter case the data may get lost...
7160 80. Added the %p formatting code to string_format() so that it can be used to
7161 print addresses in debug_print(). Adjusted all the address printing in the
7162 debugging in store.c to use %p rather than %d.
7164 81. There was a concern that a line of code in smtp_in.c could overflow a
7165 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7166 initially expressed, the concern was not well-founded, because trailing
7167 spaces are removed early. However, if the trailing spaces were followed by
7168 a NULL, they did not get removed, so the overflow was possible. Two fixes
7171 (a) I re-wrote the offending code in a cleaner fashion.
7172 (b) If an incoming SMTP command contains a NULL character, it is rejected
7175 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7176 runs initgroups(), so that if the Exim user is in any additional groups,
7177 they will be used during message reception.
7183 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob