1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
36 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
37 When redis_servers is set to a list of > 1 element, and the Redis servers
38 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
39 case of MOVED into a DEFER case instead, thus moving the query onto the
40 next server in the list. For a cluster of N elements, all N servers must
41 be defined in redis_servers.
43 GF/02 Catch and remove uninitialized value warning in exiqsumm
44 Check for existence of @ARGV before looking at $ARGV[0]
46 JH/01 Replace the store_release() internal interface with store_newblock(),
47 which internalises the check required to safely use the old one, plus
48 the allocate and data copy operations duplicated in both (!) of the
51 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
52 modifier. This matches the restriction on the commandline.
54 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
55 Previously only the last row was returned.
57 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
58 we assumed that tags in the header were well-formed, and parsed the
59 element content after inspecting only the first char of the tag.
60 Assumptions at that stage could crash the receive process on malformed
63 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
64 While running the DKIM ACL we operate on the Permanent memory pool so that
65 variables created with "set" persist to the DATA ACL. Also (at any time)
66 DNS lookups that fail create cache records using the Permanent pool. But
67 expansions release any allocations made on the current pool - so a dnsdb
68 lookup expansion done in the DKIM ACL releases the memory used for the
69 DNS negative-cache, and bad things result. Solution is to switch to the
70 Main pool for expansions.
71 While we're in that code, add checks on the DNS cache during store_reset,
72 active in the testsuite.
73 Problem spotted, and debugging aided, by Wolfgang Breyha.
75 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
76 When none of the hosts presented to a transport match an already-open
77 connection, close it and proceed with the list. Previously we would
78 queue the message. Spotted by Lena with Yahoo, probably involving
81 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
82 Previously a spurious "250 OK id=" response was appended to the proper
85 JH/08 The "support for" informational output now, which built with Content
86 Scanning support, has a line for the malware scanner interfaces compiled
87 in. Interface can be individually included or not at build time.
89 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
90 by the template makefile "src/EDITME". The "STREAM" support for an older
91 ClamAV interface method is removed.
93 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
94 rows affected is given instead).
96 JH/11 The runtime Berkeley DB library version is now additionally output by
97 "exim -d -bV". Previously only the compile-time version was shown.
99 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
100 SMTP connection. Previously, when one had more receipients than the
101 first, an abortive onward connection was made. Move to full support for
102 multiple onward connections in sequence, handling cutthrough connection
103 for all multi-message initiating connections.
105 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
106 routers. Previously, a multi-recipient message would fail to match the
107 onward-connection opened for the first recipient, and cause its closure.
109 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
110 a timeout on read on a GnuTLS initiating connection, resulting in the
111 initiating connection being dropped. This mattered most when the callout
112 was marked defer_ok. Fix to keep the two timeout-detection methods
115 JH/15 Relax results from ACL control request to enable cutthrough, in
116 unsupported situations, from error to silently (except under debug)
117 ignoring. This covers use with PRDR, frozen messages, queue-only and
120 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
122 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
123 metadata, resulting in a crash in free().
125 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
126 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
127 Broken also in d185889f4, with init system revamp.
129 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
130 Previously we abruptly closed the connection after reading a malware-
131 found indication; now we go on to read the "scan ok" response line,
134 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
135 ACL. Previously, a crash would result.
137 JH/19 Speed up macro lookups during configuration file read, by skipping non-
138 macro text after a replacement (previously it was only once per line) and
139 by skipping builtin macros when searching for an uppercase lead character.
141 JH/20 DANE support moved from Experimental to mainline. The Makefile control
142 for the build is renamed.
144 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
145 was allocated for every new TLS startup, meaning one per message. Fix
146 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
148 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
149 reported the original. Fix to report (as far as possible) the ACL
150 result replacing the original.
152 JH/23 Fix memory leak during multi-message connections using STARTTLS under
153 OpenSSL. Certificate information is loaded for every new TLS startup,
154 and the resources needed to be freed.
156 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
158 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
161 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
162 DATA response info to the (existing) per-recipient response info for
163 the "C=" log element. It can have useful tracking info from the
164 destination system. Patch from Simon Arlott.
166 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
167 length value. Previously this would segfault.
169 HS/02 Support Avast multiline protoocol, this allows passing flags to
170 newer versions of the scanner.
172 JH/28 Ensure that variables possibly set during message acceptance are marked
173 dead before release of memory in the daemon loop. This stops complaints
174 about them when the debug_store option is enabled. Discovered specifically
175 for sender_rate_period, but applies to a whole set of variables.
176 Do the same for the queue-runner and queue-list loops, for variables set
177 from spool message files. Do the same for the SMTP per-message loop, for
178 certain variables indirectly set in ACL operations.
180 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
181 as a multi-recipient message from a mailinglist manager). The coding had
182 an arbitrary cutoff number of characters while checking for more input;
183 enforced by writing a NUL into the buffer. This corrupted long / fast
184 input. The problem was exposed more widely when more pipelineing of SMTP
185 responses was introduced, and one Exim system was feeding another.
186 The symptom is log complaints of SMTP syntax error (NUL chars) on the
187 receiving system, and refused recipients seen by the sending system
188 (propating to people being dropped from mailing lists).
189 Discovered and pinpointed by David Carter.
191 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
192 replaced by the ${authresults } expansion.
194 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
196 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
197 allows proper process termination in container environments.
199 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
200 Previously the "final dot" had a newline after it; ensure it is CR,LF.
202 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
203 and "err_perm", deprecated since 4.83 when the RFC-defined words
204 "temperror" and "permerror" were introduced.
206 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
207 transport-filters or DKIM-signing. The restriction was lost in the
208 consolidation of verify-callout and delivery SMTP handling.
209 Extend the restriction to also cover ARC-signing.
211 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
212 in defer=pass mode supply a 450 to the initiator. Previously the message
215 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
216 tls_require_ciphers is used as before.
218 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
219 "pass_unscanned". Only tmpfails from the scanner are written to
220 the paniclog, as they may require admin intervention (permission
221 denied, license issues). Other scanner errors (like decompression
222 bombs) do not cause a paniclog entry.
224 JH/36 Fix reinitialisation of DKIM logging variable between messages.
225 Previously it was possible to log spurious information in receive log
228 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
229 triggered odd behaviour from Outlook Express clients.
231 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
234 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
235 since the IETF WG has not yet settled on that versus the original
236 "bare" representation.
238 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
239 Previously the millisecond value corrupted the output.
240 Fix also for syslog_pid=no and log_selector +pid, for which the pid
241 corrupted the output.
247 JH/01 Rework error string handling in TLS interface so that the caller in
248 more cases is responsible for logging. This permits library-sourced
249 string to be attached to addresses during delivery, and collapses
250 pairs of long lines into single ones.
252 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
253 during configuration. Wildcards are allowed and expanded.
255 JH/02 Rework error string handling in DKIM to pass more info back to callers.
256 This permits better logging.
258 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
259 do not close it down and have the child transport start it up again on
260 the passed-on TCP connection. Instead, proxy the child (and any
261 subsequent ones) for TLS via a unix-domain socket channel. Logging is
262 affected: the continued delivery log lines do not have any DNSSEC, TLS
263 Certificate or OCSP information. TLS cipher information is still logged.
265 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
266 identical IP addresses on different listening ports. Will also affect
269 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
270 add noisy ifdef guards to special-case this sillyness.
271 Patch from Bernd Kuhls.
273 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
274 than 255 are no longer allowed.
276 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
277 Disable the session-cache too, which might reduce our load. Since we
278 currrectly use a new context for every connection, both as server and
279 client, there is no benefit for these.
280 GnuTLS appears to not support tickets server-side by default (we don't
281 call gnutls_session_ticket_enable_server()) but client side is enabled
282 by default on recent versions (3.1.3 +) unless the PFS priority string
285 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
286 <https://reproducible-builds.org/specs/source-date-epoch/>.
288 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
289 the check for any unsuccessful recipients did not notice the limit, and
290 erroneously found still-pending ones.
292 JH/08 Pipeline CHUNKING command and data together, on kernels that support
293 MSG_MORE. Only in-clear (not on TLS connections).
295 JH/09 Avoid using a temporary file during transport using dkim. Unless a
296 transport-filter is involved we can buffer the headers in memory for
297 creating the signature, and read the spool data file once for the
298 signature and again for transmission.
300 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
301 4.77 as the kernel support then wasn't solid, having issues in 64bit
302 mode. Now, it's been long enough. Add support for FreeBSD also.
304 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
305 case where the routing stage had gathered several addresses to send to
306 a host before calling the transport for the first, we previously failed
307 to close down TLS in the old transport process before passing the TCP
308 connection to the new process. The new one sent a STARTTLS command
309 which naturally failed, giving a failed delivery and bloating the retry
310 database. Investigation and fix prototype from Wolfgang Breyha.
312 JH/12 Fix check on SMTP command input synchronisation. Previously there were
313 false-negatives in the check that the sender had not preempted a response
314 or prompt from Exim (running as a server), due to that code's lack of
315 awareness of the SMTP input buffering.
317 PP/04 Add commandline_checks_require_admin option.
318 Exim drops privileges sanely, various checks such as -be aren't a
319 security problem, as long as you trust local users with access to their
320 own account. When invoked by services which pass untrusted data to
321 Exim, this might be an issue. Set this option in main configuration
322 AND make fixes to the calling application, such as using `--` to stop
325 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
326 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
327 responses to those, into a single TLS record each way (this usually means
328 a single packet). As a side issue, smtp_enforce_sync now works on TLS
331 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
332 affects you only if you're dancing at the edge of the param size limits.
333 If you are, and this message makes sense to you, then: raise the
334 configured limit or use OpenSSL 1.1. Nothing we can do for older
337 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
338 cmdline element to get the documented default one. Previously it was
341 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
342 are unsuable for non-message-specific reasons - eg. connection timeout,
343 banner-time rejection.
345 JH/16 Fix logging of delivery remote port, when specified by router, under
348 PP/06 Repair manualroute's ability to take options in any order, even if one
349 is the name of a transport.
352 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
354 JH/17 Change the list-building routines interface to use the expanding-string
355 triplet model, for better allocation and copying behaviour.
357 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
358 Previously it was constructed the first time a possibly-matching string
359 was met in the configuration file input during startup; now it is done
362 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
363 compatible one, to avoid the (poorly documented) possibility of a config
364 file in the working directory redirecting the DB files, possibly correpting
365 some existing file. CVE-2017-10140 assigned for BDB.
367 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
368 cache-hot. Previously, although the result was properly cached, the
369 initial verify call returned a defer.
371 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
372 the main verify for receipient in uncached-mode.
374 JH/22 Retire historical build files to an "unsupported" subdir. These are
375 defined as "ones for which we have no current evidence of testing".
377 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
378 if present. Previously it was ignored.
380 JH/24 Start using specified-initialisers in C structure init coding. This is
381 a C99 feature (it's 2017, so now considered safe).
383 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
384 if was a fixed-sized field and bitmask ops via macros; it is now more
387 PP/07 GitHub PR 56: Apply MariaDB build fix.
388 Patch provided by Jaroslav Škarvada.
390 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
391 during Coverity cleanups [4.87 JH/47]
392 Diagnosis and fix provided by Michael Fischer v. Mollard.
394 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
395 the right size to place the terminating semicolon on its own folded
396 line, the header hash was calculated to an incorrect value thanks to
397 the (relaxed) space the fold became.
399 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
400 and confused the parent.
402 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
403 which could crash as a result. This could lead to undeliverable messages.
405 JH/28 Logging: "next input sent too soon" now shows where input was truncated
408 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
409 matters on fast-turnover and PID-randomising systems, which were getting
410 out-of-order delivery.
412 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
413 a possibly-overlapping copy. The symptom was that "Remote host closed
414 connection in response to HELO" was logged instead of the actual 4xx
417 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
418 Previously only that bufferd was discarded, resulting in SYMTP command
421 JH/32 DKIM: when a message has multiple signatures matching an identity given
422 in dkim_verify_signers, run the dkim acl once for each. Previously only
423 one run was done. Bug 2189.
425 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
426 "panic the current process" to "deliberately defer". The panic log is
427 still written with the problem list name; the mail and reject logs now
428 get a temp-reject line for the message that was being handled, saying
429 something like "domains check lookup or other defer". The SMTP 451
430 message is still "Temporary local problem".
432 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
433 A crafted sequence of BDAT commands could result in in-use memory beeing
434 freed. CVE-2017-16943.
436 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
437 from SMTP input. Previously it was always done; now only done for DATA
438 and not BDAT commands. CVE-2017-16944.
440 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
441 to the message (such as an overlong header line). Previously this was
442 not done and we did not exit BDAT mode. Followon from the previous item
443 though a different problem.
449 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
450 than -2003 did; needs libidn2 in addition to libidn.
452 JH/02 The path option on a pipe transport is now expanded before use.
454 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
455 Patch provided by "Björn", documentation fix added too.
457 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
458 missing a wire-to-host endian conversion.
460 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
461 close after a BDAT command line could be taken as a following command,
462 giving a synch failure. Fix by only checking for synch immediately
463 before acknowledging the chunk.
465 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
466 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
467 macro. Patches provided by Josh Soref.
469 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
470 Previously we did not; the RFC seems ambiguous and VRFY is not listed
471 by IANA as a service extension. However, John Klensin suggests that we
474 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
475 the dkim code may be unix-mode line endings rather than smtp wire-format
476 CRLF, so prepend a CR to any bare LF.
478 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
479 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
481 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
482 the first were themselves being wrongly included in the feed into dkim
483 processing; with most chunk sizes in use this resulted in an incorrect
484 body hash calculated value.
486 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
487 DKIM signature block, for verification. Although advised against by
488 standards it is specifically not ruled illegal.
490 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
492 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
493 missing a body hash (the bh= tag).
495 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
496 It seems that HAProxy sends the Proxy Protocol information in clear and
497 only then does a TLS startup, so do the same.
499 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
500 TCP connections (such as for Spamd) unless the daemon successfully set
501 Fast Open mode on its listening sockets. This fixes breakage seen on
502 too-old kernels or those not configured for Fast Open, at the cost of
503 requiring both directions being enabled for TFO, and TFO never being used
504 by non-daemon-related Exim processes.
506 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
507 endings, at least on the first header line. Try to canonify any that get
508 past that check, despite the cost.
510 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
511 now limited to an arbitrary five deep, while parsing addresses with the
512 strip_excess_angle_brackets option enabled.
514 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
515 instead leave the unprompted TLS handshake in socket buffer for the
516 TLS library to consume.
518 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
520 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
522 JH/16 Drop variables when they go out of scope. Memory management drops a whole
523 region in one operation, for speed, and this leaves assigned pointers
524 dangling. Add checks run only under the testsuite which checks all
525 variables at a store-reset and panics on a dangling pointer; add code
526 explicitly nulling out all the variables discovered. Fixes one known
527 bug: a transport crash, where a dangling pointer for $sending_ip_address
528 originally assigned in a verify callout, is re-used.
530 PP/06 Drop '.' from @INC in various Perl scripts.
532 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
534 PP/08 Reduce a number of compilation warnings under clang; building with
535 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
536 should be warning-free.
538 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
540 HS/01 Fix portability problems introduced by PP/08 for platforms where
541 realloc(NULL) is not equivalent to malloc() [SunOS et al].
543 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
544 chunk. This allows us to accept broken chunked messages. We need a more
545 general solution here.
547 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
548 already-broken messages in the queue.
550 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
552 JH/19 Fix reference counting bug in routing-generated-address tracking.
558 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
559 supports it and a size is available (ie. the sending peer gave us one).
561 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
562 years of being deprecated). The replacements are the ACLs
563 acl_smtp_mime and acl_not_smtp_mime.
565 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
566 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
567 or even an in-clear connection were permitted. Now, if the host lookup
568 was dnssec and dane was requested then the host is only used if the
569 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
570 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
571 if one fails this test.
572 This means that a poorly-configured remote DNS will make it incommunicado;
573 but it protects against a DNS-interception attack on it.
575 JH/04 Bug 1810: make continued-use of an open smtp transport connection
576 non-noisy when a race steals the message being considered.
578 JH/05 If main configuration option tls_certificate is unset, generate a
579 self-signed certificate for inbound TLS connections.
581 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
582 in rewrites and routers.
584 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
585 and logged a warning sing 4.83; now they are a configuration file error.
587 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
588 (lacking @domain). Apply the same qualification processing as RCPT.
590 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
592 JH/10 Support ${sha256:} applied to a string (as well as the previous
595 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
596 a cutthrough deliver is pending, as we always want to make a connection.
597 This also avoids re-routing the message when later placing the cutthrough
598 connection after a verify cache hit.
599 Do not update it with the verify result either.
601 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
602 when routing results in more than one destination address.
604 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
605 signing (which inhibits the cutthrough capability). Previously only
606 the presence of an option was tested; now an expansion evaluating as
607 empty is permissible (obviously it should depend only on data available
608 when the cutthrough connection is made).
610 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
611 the relevant preceding SMTP command did not note the pipelining mode.
613 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
614 Previously they were not counted.
616 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
617 as one having no matching records. Previously we deferred the message
618 that needed the lookup.
620 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
621 distinguished as "(=".
623 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
624 for missing MX records. Previously it only worked for missing A records.
626 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
628 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
629 after the data-go-ahead and data-ack. Patch from Jason Betts.
631 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
632 even for a "none" policy. Patch from Tony Meyer.
634 JH/22 Fix continued use of a connection for further deliveries. If a port was
635 specified by a router, it must also match for the delivery to be
638 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
639 When one of the recipients of a message was unsuitable for the connection
640 (has no matching addresses), we lost track of needing to mark it
641 deferred. As a result mail would be lost.
643 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
645 JH/25 Decoding ACL controls is now done using a binary search; the source code
646 takes up less space and should be simpler to maintain. Merge the ACL
647 condition decode tables also, with similar effect.
649 JH/26 Fix problem with one_time used on a redirect router which returned the
650 parent address unchanged. A retry would see the parent address marked as
651 delivered, so not attempt the (identical) child. As a result mail would
654 JH/27 Fix a possible security hole, wherein a process operating with the Exim
655 UID can gain a root shell. Credit to http://www.halfdog.net/ for
656 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
659 JH/28 Enable {spool,log} filesystem space and inode checks as default.
660 Main config options check_{log,spool}_{inodes,space} are now
661 100 inodes, 10MB unless set otherwise in the configuration.
663 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
664 Previously it only applied to the main-section connection policy
667 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
669 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
670 by me. Added RFC7919 DH primes as an alternative.
672 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
673 are not in the system include path.
675 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
676 GnuTLS, when a session startup failed (eg because the client disconnected)
677 Exim did stdio operations after fclose. This was exposed by a recent
678 change which nulled out the file handle after the fclose.
680 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
681 signed directly by the cert-signing cert, rather than an intermediate
682 OCSP-signing cert. This is the model used by LetsEncrypt.
684 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
686 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
687 an incoming connection.
689 HS/02 Bug 1802: Do not half-close the connection after sending a request
692 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
693 fallback to "prime256v1".
695 JH/34 SECURITY: Use proper copy of DATA command in error message.
696 Could leak key material. Remotely exploitable. CVE-2016-9963.
702 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
703 and 3.4.4 - once the server is enabled to respond to an OCSP request
704 it does even when not requested, resulting in a stapling non-aware
705 client dropping the TLS connection.
707 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
708 support variable-length bit vectors. No functional change.
710 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
711 The I= interface field on outgoing lines is now after the H= remote
712 host field, same as incoming lines. There is a separate
713 outgoing_interface log selector which allows you to disable the
716 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
717 If not running log_selector +smtp_connection the mainlog would be held
718 open indefinitely after a "too many connections" event, including to a
719 deleted file after a log rotate. Leave the per net connection logging
720 leaving it open for efficiency as that will be quickly detected by the
721 check on the next write.
723 HS/01 Bug 1671: Fix post transport crash.
724 Processing the wait-<transport> messages could crash the delivery
725 process if the message IDs didn't exist for some reason. When
726 using 'split_spool_directory=yes' the construction of the spool
727 file name failed already, exposing the same netto behaviour.
729 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
730 mime_regex ACL conditions.
732 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
733 to DSN fail messages (bounces): remote IP, remote greeting, remote response
734 to HELO, local diagnostic string.
736 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
737 log line to debug. Even when configured with a tls authenticator many
738 client connections are expected to not authenticate in this way, so
739 an authenticate fail is not an error.
741 HS/02 Add the Exim version string to the process info. This way exiwhat
742 gives some more detail about the running daemon.
744 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
745 matter for fast-change records such as DNSBLs.
747 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
748 retry record, even if constant. There may be multiple transports with
749 different interface settings and the retry behaviour needs to be kept
752 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
754 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
756 JH/10 Bug 840: fix log_defer_output option of pipe transport
758 JH/11 Bug 830: use same host for all RCPTS of a message, even under
759 hosts_randomize. This matters a lot when combined with mua_wrapper.
761 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
762 ${quote_pgsql:<string>} operator.
764 JH/13 Bug 1708: avoid misaligned access in cached lookup.
766 JH/14 Change header file name for freeradius-client. Relevant if compiling
767 with Radius support; from the Gentoo tree and checked under Fedora.
769 JH/15 Bug 1712: Introduce $prdr_requested flag variable
771 JH/16 Bug 1714: Permit an empty string as expansion result for transport
772 option transport_filter, meaning no filtering.
774 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
776 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
777 defaults to "*" (all hosts). The variable is now available when not built
778 with TLS, default unset, mainly to enable keeping the testsuite sane.
779 If a server certificate is not supplied (via tls_certificate) an error is
780 logged, and clients will find TLS connections fail on startup. Presumably
781 they will retry in-clear.
782 Packagers of Exim are strongly encouraged to create a server certificate
783 at installation time.
785 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
786 with the $config_file variable.
788 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
789 in transport context, after the attempt, and per-recipient. The latter type
790 is per host attempted. The event data is the error message, and the errno
791 information encodes the lookup type (A vs. MX) used for the (first) host,
792 and the trailing two digits of the smtp 4xx response.
794 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
795 to write to mainlog (or rejectlog, paniclog) in the window between file
796 creation and permissions/ownership being changed. Particularly affects
797 installations where exicyclog is run as root, rather than exim user;
798 result is that the running daemon panics and dies.
800 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
802 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
803 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
804 "pri" and "weight". Note that the previous implicit priority given by the
805 list order is no longer honoured.
807 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
810 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
811 by defining SUPPORT_SOCKS.
813 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
814 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
815 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
816 variables are renamed to proxy_{local,external}_{address,port}.
818 JH/27 Move Internationalisation support from Experimental to mainline, enabled
819 for a build by defining SUPPORT_I18N
821 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
822 of the query string, and make ${quote_redis:} do that quoting.
824 JH/29 Move Events support from Experimental to mainline, enabled by default
825 and removable for a build by defining DISABLE_EVENT.
827 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
829 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
830 cached by the daemon.
832 JH/32 Move Redis support from Experimental to mainline, enabled for a build
833 by defining LOOKUP_REDIS. The libhiredis library is required.
835 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
836 keys are given for lookup.
838 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
839 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
840 only supported when built with TLS support. The PolarSSL SHA routines
841 are still used when the TLS library is too old for convenient support.
843 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
844 openssl_options), for security. OpenSSL forces this from version 1.1.0
845 server-side so match that on older versions.
847 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
848 allocation for $value could be released as the expansion processing
849 concluded, but leaving the global pointer active for it.
851 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
852 and to use the domains and local_parts ACL conditions.
854 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
855 incorrectly not doubled on cutthrough transmission, hence seen as a
856 body-termination at the receiving system - resulting in truncated mails.
857 Commonly the sender saw a TCP-level error, and retransmitted the message
858 via the normal store-and-forward channel. This could result in duplicates
859 received - but deduplicating mailstores were liable to retain only the
860 initial truncated version.
862 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
864 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
866 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
867 we're in there, support oversigning also; bug 1309.
869 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
871 HS/04 Add support for keep_environment and add_environment options.
873 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
874 either intentional arithmetic overflow during PRNG, or testing config-
877 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
878 delivery resulted in actual delivery. Cancel cutthrough before DATA
881 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
882 rejects a recipient: pass the reject to the originator.
884 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
885 Many were false-positives and ignorable, but it's worth fixing the
888 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
889 for the new environment-manipulation done at startup. Move the routines
890 from being local to tls.c to being global via the os.c file.
892 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
893 an extract embedded as result-arg for a map, the first arg for extract
894 is unavailable so we cannot tell if this is a numbered or keyed
895 extraction. Accept either.
901 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
904 JH/02 The smtp transport option "multi_domain" is now expanded.
906 JH/03 The smtp transport now requests PRDR by default, if the server offers
909 JH/04 Certificate name checking on server certificates, when exim is a client,
910 is now done by default. The transport option tls_verify_cert_hostnames
911 can be used to disable this per-host. The build option
912 EXPERIMENTAL_CERTNAMES is withdrawn.
914 JH/05 The value of the tls_verify_certificates smtp transport and main options
915 default to the word "system" to access the system default CA bundle.
916 For GnuTLS, only version 3.0.20 or later.
918 JH/06 Verification of the server certificate for a TLS connection is now tried
919 (but not required) by default. The verification status is now logged by
920 default, for both outbound TLS and client-certificate supplying inbound
923 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
926 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
927 Status Notification (bounce) messages are now MIME format per RFC 3464.
928 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
929 under the control of the dsn_advertise_hosts option, and routers may
930 have a dsn_lasthop option.
932 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
933 default, modifiable by a malware= option. The list separator for
934 the options can now be changed in the usual way. Bug 68.
936 JH/10 The smtp_receive_timeout main option is now expanded before use.
938 JH/11 The incoming_interface log option now also enables logging of the
939 local interface on delivery outgoing connections.
941 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
942 if the interface and destination host and port all match.
944 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
947 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
948 Patch from Andrew Lewis.
950 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
951 now supports optional time-restrictions, weighting, and priority
952 modifiers per server. Patch originally by <rommer@active.by>.
954 JH/16 The spamd_address main option now supports a mixed list of local
955 and remote servers. Remote servers can be IPv6 addresses, and
956 specify a port-range.
958 JH/17 Bug 68: The spamd_address main option now supports an optional
959 timeout value per server.
961 JH/18 Bug 1581: Router and transport options headers_add/remove can
962 now have the list separator specified.
964 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
967 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
970 JH/21 Support for the A6 type of dns record is withdrawn.
972 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
973 rather than the verbs used.
975 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
976 from 255 to 1024 chars.
978 JH/24 Verification callouts now attempt to use TLS by default.
980 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
981 are generic router options now. The defaults didn't change.
983 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
984 Original patch from Alexander Shikoff, worked over by JH.
986 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
987 files if ZCAT_COMMAND is not executable.
989 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
991 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
993 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
994 Normally benign, it bites when the pair was led to by a CNAME;
995 modern usage is to not canonicalize the domain to a CNAME target
996 (and we were inconsistent anyway for A-only vs AAAA+A).
998 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1000 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1001 when evaluating $sender_host_dnssec.
1003 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1004 $sender_helo_dnssec variable.
1006 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1008 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1010 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1012 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1013 documented as working, but never had. Support all but $spam_report.
1015 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1016 added for tls authenticator.
1018 HS/03 Add perl_taintmode main config option
1024 TL/01 When running the test suite, the README says that variables such as
1025 no_msglog_check are global and can be placed anywhere in a specific
1026 test's script, however it was observed that placement needed to be near
1027 the beginning for it to behave that way. Changed the runtest perl
1028 script to read through the entire script once to detect and set these
1029 variables, reset to the beginning of the script, and then run through
1030 the script parsing/test process like normal.
1032 TL/02 The BSD's have an arc4random API. One of the functions to induce
1033 adding randomness was arc4random_stir(), but it has been removed in
1034 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1035 function when detected.
1037 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1038 cause callback expansion.
1040 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1041 syntax errors in an expansion can be treated as a string instead of
1042 logging or causing an error, due to the internal use of bool_lax
1043 instead of bool when processing it.
1045 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1046 server certificates when making smtp deliveries.
1048 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1050 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1052 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1054 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1055 Merged patch from Sebastian Wiedenroth.
1057 JH/05 Fix results-pipe from transport process. Several recipients, combined
1058 with certificate use, exposed issues where response data items split
1059 over buffer boundaries were not parsed properly. This eventually
1060 resulted in duplicates being sent. This issue only became common enough
1061 to notice due to the introduction of connection certificate information,
1062 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1064 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1065 size buffer was used, resulting in syntax errors when an expansion
1068 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1069 version 3.3.6 or later.
1071 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1072 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1073 both become "event_action", the variables become $event_name, $event_data
1074 and $event_defer_errno. There is a new variable $verify_mode, usable in
1075 routers, transports and related events. The tls:cert event is now also
1076 raised for inbound connections, if the main configuration event_action
1079 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1080 early OCSP support, but no stapling (appears to be less than 1.0.0).
1082 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1083 server certificate names available under the smtp transport option
1084 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1087 JH/10 Time-related extraction expansions from certificates now use the main
1088 option "timezone" setting for output formatting, and are consistent
1089 between OpenSSL and GnuTLS compilations. Bug 1541.
1091 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1092 encoded parameter in the incoming message. Bug 1558.
1094 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1095 include certificate info, eximon was claiming there were spoolfile
1098 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1100 JH/14 Log delivery-related information more consistently, using the sequence
1101 "H=<name> [<ip>]" wherever possible.
1103 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1104 are problematic for Debian distribution, omit them from the release
1107 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1109 JH/16 Fix string representation of time values on 64bit time_t architectures.
1112 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1113 output list separator was used.
1118 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1119 checkers that were complaining about end of non-void function with no
1122 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1123 This was a regression introduced in 4.83 by another bugfix.
1125 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1127 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1128 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1134 TF/01 Correctly close the server side of TLS when forking for delivery.
1136 When a message was received over SMTP with TLS, Exim failed to clear up
1137 the incoming connection properly after forking off the child process to
1138 deliver the message. In some situations the subsequent outgoing
1139 delivery connection happened to have the same fd number as the incoming
1140 connection previously had. Exim would try to use TLS and fail, logging
1141 a "Bad file descriptor" error.
1143 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1144 utilities have not been installed.
1146 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1147 temporary space as the ACL may create new global variables.
1149 TL/01 LDAP support uses per connection or global context settings, depending
1150 upon the detected version of the libraries at build time.
1152 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1153 to extract and use the src ip:port in logging and expansions as if it
1154 were a direct connection from the outside internet. PPv2 support was
1155 updated based on HAProxy spec change in May 2014.
1157 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1159 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1160 Properly escape header and check for NULL return.
1162 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1165 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1167 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1168 characters in header names, implemented as a verify condition.
1169 Contributed by Michael Fischer v. Mollard.
1171 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1172 results permerror and temperror. Previous values are deprecated but
1173 still accepted. In a future release, err_perm and err_temp will be
1174 completely removed, which will be a backward incompatibility if the
1175 ACL tests for either of these two old results. Patch contributed by
1176 user bes-internal on the mailing list.
1178 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1180 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1181 selectors, in both main and reject logs.
1183 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1186 JH/07 Add malware type "sock" for talking to simple daemon.
1188 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1190 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1191 routers/transports under cutthrough routing.
1193 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1194 numbers. Touch up "bool" conditional to keep the same definition.
1196 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1198 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1200 JH/12 Expand items in router/transport headers_add or headers_remove lists
1201 individually rather than the list as a whole. Bug 1452.
1203 Required for reasonable handling of multiple headers_ options when
1204 they may be empty; requires that headers_remove items with embedded
1205 colons must have them doubled (or the list-separator changed).
1207 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1208 view the policy declared in the DMARC record. Currently, $dmarc_status
1209 is a combined value of both the record presence and the result of the
1212 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1214 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1215 dnslookup router and the smtp transport (applying to the forward
1218 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1219 of ldap servers used for a specific lookup. Patch provided by Heiko
1222 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1223 New variable $lookup_dnssec_authenticated for observability.
1225 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1226 Patch submitted by Lars Timman.
1228 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1230 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1231 Requires trusted mode and valid format message id, aborts otherwise.
1232 Patch contributed by Heiko Schlichting.
1234 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1235 certextract with support for various fields. Bug 1358.
1237 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1238 is requested by default, modifiable by smtp transport option
1241 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1242 operate on certificate variables to give certificate fingerprints
1243 Also new ${sha256:cert_variable}.
1245 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1247 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1248 Christian Aistleitner.
1250 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1252 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1253 file. Patch from Wolfgang Breyha.
1255 JH/25 Expand the coverage of the delivery $host and $host_address to
1256 client authenticators run in verify callout. Bug 1476.
1258 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1259 align with daemon_smtp_ports. Bug 72.
1261 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1262 support and error reporting did not work properly.
1264 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1265 and is readable. Patch from Andrew Colin Kissa.
1267 TL/14 Enhance documentation of ${run expansion and how it parses the
1268 commandline after expansion, particularly in the case when an
1269 unquoted variable expansion results in an empty value.
1271 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1273 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1274 by renaming away from it. Addresses GH issue 15, reported by
1277 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1279 TL/15 SECURITY: prevent double expansion in math comparison functions
1280 (can expand unsanitized data). Not remotely exploitable.
1287 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1289 PP/02 Make -n do something, by making it not do something.
1290 When combined with -bP, the name of an option is not output.
1292 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1295 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1296 $sender_host_name and config options to manage this, and basic check
1299 PP/05 DSCP support for outbound connections and control modifier for inbound.
1301 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1302 (Only plugin which currently uses this is kerberos4, which nobody should
1303 be using, but we should make it available and other future plugins might
1304 conceivably use it, even though it would break NAT; stuff *should* be
1305 using channel bindings instead).
1307 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1308 name; added for Sendmail compatibility; requires admin caller.
1309 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1310 just ignore it); requires trusted caller.
1311 Also parse but ignore: -Ac -Am -X<logfile>
1314 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1316 TL/02 Add +smtp_confirmation as a default logging option.
1318 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1319 Patch by Magnus Holmgren from 2007-02-20.
1321 TL/04 Bugzilla 1281 - Spec typo.
1322 Bugzilla 1283 - Spec typo.
1323 Bugzilla 1290 - Spec grammar fixes.
1325 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1327 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1329 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1330 mailing list by by Dmitry Isaikin.
1332 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1334 JH/02 Support "G" suffix to numbers in ${if comparisons.
1336 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1338 NM/01 Bugzilla 1197 - Spec typo
1339 Bugzilla 1196 - Spec examples corrections
1341 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1343 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1344 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1347 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1348 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1350 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1351 "acl {{name}{arg}...}", and optional args on acl condition
1354 JH/05 Permit multiple router/transport headers_add/remove lines.
1356 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1358 JH/07 Avoid using a waiting database for a single-message-only transport.
1359 Performance patch from Paul Fisher. Bugzilla 1262.
1361 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1364 JH/09 Add $headers_added variable, with content from use of ACL modifier
1365 add_header (but not yet added to the message). Bugzilla 199.
1367 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1368 Pulled from Bugzilla 817 by Wolfgang Breyha.
1370 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1372 (nb: this is the same fix as in Exim 4.80.1)
1374 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1377 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1380 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1382 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1383 advertise SMTP AUTH mechanism to us, instead of a generic
1384 protocol violation error. Also, make Exim more robust to bad
1385 data from the Dovecot auth socket.
1387 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1389 When a queue runner is handling a message, Exim first routes the
1390 recipient addresses, during which it prunes them based on the retry
1391 hints database. After that it attempts to deliver the message to
1392 any remaining recipients. It then updates the hints database using
1395 So if a recipient address works intermittently, it can get repeatedly
1396 deferred at routing time. The retry hints record remains fresh so the
1397 address never reaches the final cutoff time.
1399 This is a fairly common occurrence when a user is bumping up against
1400 their storage quota. Exim had some logic in its local delivery code
1401 to deal with this. However it did not apply to per-recipient defers
1402 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1404 This change adds a proper retry rule check during routing so that the
1405 final cutoff time is checked against the message's age. We only do
1406 this check if there is an address retry record and there is not a
1407 domain retry record; this implies that previous attempts to handle
1408 the address had the retry_use_local_parts option turned on. We use
1409 this as an approximation for the destination being like a local
1410 delivery, as in LMTP.
1412 I suspect this new check makes the old local delivery cutoff check
1413 redundant, but I have not verified this so I left the code in place.
1415 TF/02 Correct gecos expansion when From: is a prefix of the username.
1417 Test 0254 submits a message to Exim with the header
1421 When I ran the test suite under the user fanf2, Exim expanded
1422 the header to contain my full name, whereas it should have added
1423 a Resent-Sender: header. It erroneously treats any prefix of the
1424 username as equal to the username.
1426 This change corrects that bug.
1428 GF/01 DCC debug and logging tidyup
1429 Error conditions log to paniclog rather than rejectlog.
1430 Debug lines prefixed by "DCC: " to remove any ambiguity.
1432 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1434 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1435 Bug spotted by Jeremy Harris; was flawed since initial commit.
1436 Would have resulted in OCSP responses post-SNI triggering an Exim
1437 NULL dereference and crash.
1439 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1441 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1442 Bug detection, analysis and fix by Samuel Thibault.
1443 Bugzilla 1331, Debian bug #698092.
1445 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1447 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1448 Server implementation by Todd Lyons, client by JH.
1449 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1450 config variable "prdr_enable" controls whether the server
1451 advertises the facility. If the client requests PRDR a new
1452 acl_data_smtp_prdr ACL is called once for each recipient, after
1453 the body content is received and before the acl_smtp_data ACL.
1454 The client is controlled by both of: a hosts_try_prdr option
1455 on the smtp transport, and the server advertisement.
1456 Default client logging of deliveries and rejections involving
1457 PRDR are flagged with the string "PRDR".
1459 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1460 fclose(). Diagnosis by Todd Lyons.
1462 PP/17 Update configure.default to handle IPv6 localhost better.
1463 Patch by Alain Williams (plus minor tweaks).
1466 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1467 This is now consistent with GnuTLS, and is now documented: the
1468 previous undocumented portable approach to treating the option as
1469 unset was to force an expansion failure. That still works, and
1470 an empty string is now equivalent.
1472 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1473 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1474 not performing validation itself.
1476 PP/20 Added force_command boolean option to pipe transport.
1477 Patch from Nick Koston, of cPanel Inc.
1479 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1482 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1484 PP/21 Fix eximon continuous updating with timestamped log-files.
1485 Broken in a format-string cleanup in 4.80, missed when I repaired the
1486 other false fix of the same issue.
1487 Report and fix from Heiko Schlichting.
1490 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1491 Report from Prashanth Katuri.
1493 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1494 It's SecureTransport, so affects any MacOS clients which use the
1495 system-integrated TLS libraries, including email clients.
1497 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1498 using a MIME ACL for non-SMTP local injection.
1499 Report and assistance in diagnosis by Warren Baker.
1501 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1503 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1505 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1506 last id that failed so it may be referenced in subsequent ACL's.
1508 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1511 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1512 ldap library initialization, allowing self-signed CA's to be
1513 used. Also properly sets require_cert option later in code by
1514 using NULL (global ldap config) instead of ldap handle (per
1515 session). Bug diagnosis and testing by alxgomz.
1517 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1518 the src/util/ subdirectory.
1520 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1521 renamed to Transport Post Delivery Action by Jeremy Harris, as
1524 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1525 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1526 redis_servers = needs to be configured which will be used by the redis
1527 lookup. Patch from Warren Baker, of The Packet Hub.
1529 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1531 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1532 hostname or reverse DNS when processing a host list. Used suggestions
1533 from multiple comments on this bug.
1535 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1537 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1538 Missed a few lines, added it to make the runtest require no keyboard
1541 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1542 contains upper case chars. Make router use caseful_local_part.
1544 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1545 support when GnuTLS has been built with p11-kit.
1551 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1553 This, or similar/improved, will also be change PP/11 of 4.82.
1559 PP/01 Handle short writes when writing local log-files.
1560 In practice, only affects FreeBSD (8 onwards).
1561 Bugzilla 1053, with thanks to Dmitry Isaikin.
1563 NM/01 Bugzilla 949 - Documentation tweak
1565 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1568 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1570 PP/02 Implemented gsasl authenticator.
1572 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1574 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1575 `pkg-config foo` for cflags/libs.
1577 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1578 with rest of GSASL and with heimdal_gssapi.
1580 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1581 `pkg-config foo` for cflags/libs for the TLS implementation.
1583 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1584 properties get this fed in as external SSF. A number of robustness
1585 and debugging improvements to the cyrus_sasl authenticator.
1587 PP/08 cyrus_sasl server now expands the server_realm option.
1589 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1590 Patch by Jeremy Harris.
1592 PP/10 Added dbmjz lookup type.
1594 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1596 PP/12 MAIL args handles TAB as well as SP, for better interop with
1597 non-compliant senders.
1598 Analysis and variant patch by Todd Lyons.
1600 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1601 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1602 Patch from Dirk Mueller <dmueller@suse.com>
1604 PP/13 tls_peerdn now print-escaped for spool files.
1605 Observed some $tls_peerdn in wild which contained \n, which resulted
1606 in spool file corruption.
1608 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1609 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1610 or write after TLS renegotiation, which otherwise led to messages
1613 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1614 as a tracking header (ie: a signed header comes before the signature).
1615 Patch from Wolfgang Breyha.
1617 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1618 comma-sep list; embedded commas doubled.
1620 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1622 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1624 Report and patch from Dmitry Banschikov.
1626 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1627 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1628 failures. We appear to now support TLS1.1+ with Exim.
1630 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1631 lets Exim select keys and certificates based upon TLS SNI from client.
1632 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1633 before an outbound SMTP session. New log_selector, +tls_sni.
1635 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1636 NULL dereference. Report and patch from Alun Jones.
1638 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1639 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1640 Not seeing resolver debug output on NetBSD, but suspect this is a
1641 resolver implementation change.
1643 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1644 Left warnings. Added "eximon gdb" invocation mode.
1646 PP/21 Defaulting "accept_8bitmime" to true, not false.
1648 PP/22 Added -bw for inetd wait mode support.
1650 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1651 locate the relevant includes and libraries. Made this the default.
1653 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1654 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1656 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1657 This may cause build issues on older platforms.
1659 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1660 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1661 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1662 Added SNI support via GnuTLS too.
1663 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1665 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1667 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1668 Applied second patch from Janne, implementing suggestion to default
1669 multiple-strings-in-record handling to match SPF spec.
1671 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1673 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1674 read-only, out of scope).
1675 Patch from Wolfgang Breyha, report from Stuart Northfield.
1677 PP/29 Fix three issues highlighted by clang analyser static analysis.
1678 Only crash-plausible issue would require the Cambridge-specific
1679 iplookup router and a misconfiguration.
1680 Report from Marcin Mirosław.
1682 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1684 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1685 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1686 As part of this, removing so much warning spew let me fix some minor
1687 real issues in debug logging.
1689 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1690 assignment on my part. Fixed.
1692 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1693 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1694 Janne Snabb (who went above and beyond: thank you).
1696 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1697 string otherwise requires a connection and a bunch more work and it's
1698 relatively easy to get wrong. Should also expose TLS library linkage
1701 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1702 64-bit ${eval} (JH/03).
1704 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1705 GNU libc to support some of the 64-bit stuff, should not lead to
1706 conflicts. Defined before os.h is pulled in, so if a given platform
1707 needs to override this, it can.
1709 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1710 protection layer was required, which is not implemented.
1711 Bugzilla 1254, patch from Wolfgang Breyha.
1713 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1714 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1715 tls_dhparam take prime identifiers. Also unbreak combination of
1716 OpenSSL+DH_params+TLSSNI.
1718 PP/39 Disable SSLv2 by default in OpenSSL support.
1724 PP/01 Solaris build fix for Oracle's LDAP libraries.
1725 Bugzilla 1109, patch from Stephen Usher.
1727 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1729 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1732 TF/02 Fix a couple more cases where we did not log the error message
1733 when unlink() failed. See also change 4.74-TF/03.
1735 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1736 lock up or crash if it happened to be inside a call to libc when it
1737 got a SIGUSR1 from exiwhat.
1739 The SIGUSR1 handler appends the current process status to the process
1740 log which is later printed by exiwhat. It used to use the general
1741 purpose logging code to do this, but several functions it calls are
1742 not safe for signals.
1744 The new output code in the SIGUSR1 handler is specific to the process
1745 log, and simple enough that it's easy to inspect for signal safety.
1746 Removing some special cases also simplifies the general logging code.
1747 Removing the spurious timestamps from the process log simplifies
1750 TF/04 Improved ratelimit ACL condition.
1752 The /noupdate option has been deprecated in favour of /readonly which
1753 has clearer semantics. The /leaky, /strict, and /readonly update modes
1754 are mutually exclusive. The update mode is no longer included in the
1755 database key; it just determines when the database is updated. (This
1756 means that when you upgrade Exim will forget old rate measurements.)
1758 Exim now checks that the per_* options are used with an update mode that
1759 makes sense for the current ACL. For example, when Exim is processing a
1760 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1761 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1762 must specify per_mail/readonly. If you omit the update mode it defaults to
1763 /leaky where that makes sense (as before) or /readonly where required.
1765 The /noupdate option is now undocumented but still supported for
1766 backwards compatibility. It is equivalent to /readonly except that in
1767 ACLs where /readonly is required you may specify /leaky/noupdate or
1768 /strict/noupdate which are treated the same as /readonly.
1770 A useful new feature is the /count= option. This is a generalization
1771 of the per_byte option, so that you can measure the throughput of other
1772 aggregate values. For example, the per_byte option is now equivalent
1773 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1775 The per_rcpt option has been generalized using the /count= mechanism
1776 (though it's more complicated than the per_byte equivalence). When it is
1777 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1778 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1779 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1780 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1781 Note that using per_rcpt with a non-readonly update mode in more than
1782 one ACL will cause the recipients to be double-counted. (The per_mail
1783 and per_byte options don't have this problem.)
1785 The handling of very low rates has changed slightly. If the computed rate
1786 is less than the event's count (usually one) then this event is the first
1787 after a long gap. In this case the rate is set to the same as this event's
1788 count, so that the first message of a spam run is counted properly.
1790 The major new feature is a mechanism for counting the rate of unique
1791 events. The new per_addr option counts the number of different
1792 recipients that someone has sent messages to in the last time period. It
1793 behaves like per_rcpt if all the recipient addresses are different, but
1794 duplicate recipient addresses do not increase the measured rate. Like
1795 the /count= option this is a general mechanism, so the per_addr option
1796 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1797 example, measure the rate that a client uses different sender addresses
1798 with the options per_mail/unique=$sender_address. There are further
1799 details in the main documentation.
1801 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1803 TF/06 Removed a few PCRE remnants.
1805 TF/07 Automatically extract Exim's version number from tags in the git
1806 repository when doing development or release builds.
1808 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1809 Bugzilla 879. Patch from Paul Fisher.
1811 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1812 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1815 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1817 PP/05 Variable $av_failed, true if the AV scanner deferred.
1818 Bugzilla 1078. Patch from John Horne.
1820 PP/06 Stop make process more reliably on build failure.
1821 Bugzilla 1087. Patch from Heiko Schlittermann.
1823 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1824 Bugzilla 1089. Patch from Heiko Schlittermann.
1826 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1827 Bugzilla 1131. Patch from Holger Weiß.
1829 PP/09 Handle IPv6 addresses with SPF.
1830 Bugzilla 860. Patch from Wolfgang Breyha.
1832 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1834 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1837 PP/11 match_* no longer expand right-hand-side by default.
1838 New compile-time build option, EXPAND_LISTMATCH_RHS.
1839 New expansion conditions, "inlist", "inlisti".
1841 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1843 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1845 PP/14 fix log_write() format string regression from TF/03.
1846 Bugzilla 1152. Patch from Dmitry Isaikin.
1852 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1854 PP/02 Harmonised TLS library version reporting; only show if debugging.
1855 Layout now matches that introduced for other libraries in 4.74 PP/03.
1857 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1859 PP/04 New "dns_use_edns0" global option.
1861 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1864 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1865 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1867 TK/01 Updated PolarSSL code to 0.14.2.
1868 Bugzilla 1097. Patch from Andreas Metzler.
1870 PP/07 Catch divide-by-zero in ${eval:...}.
1871 Fixes bugzilla 1102.
1873 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1876 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1877 format-string attack -- SECURITY: remote arbitrary code execution.
1879 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1880 time unintentionally subject to list matching rules, letting the header
1881 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1882 arbitrary string expansion). This allowed for information disclosure.
1884 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1885 INT_MIN/-1 -- value coerced to INT_MAX.
1891 NM/01 Workaround for PCRE version dependency in version reporting
1894 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1895 This fixes portability to compilers other than gcc, notably
1896 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1898 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1899 makefiles for portability to HP-UX and POSIX correctness.
1901 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1902 Also via indented variable definition in the Makefile.
1903 (Debugging by Oliver Heesakkers).
1905 PP/02 Restore caching of spamd results with expanded spamd_address.
1906 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1908 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1909 Improves build reliability. Fix from: Frank Elsner
1911 NM/02 Fix wide character breakage in the rfc2047 coding
1912 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1914 NM/03 Allow underscore in dnslist lookups
1915 Fixes bug 1026. Patch from Graeme Fowler
1917 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1918 Code patches from Adam Ciarcinski of NetBSD.
1920 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1923 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1924 is logged, to avoid truncation. Patch from John Horne.
1926 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1927 Patch from Jakob Hirsch.
1929 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1930 SQL string expansion failure details.
1931 Patch from Andrey Oktyabrski.
1933 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1934 Patch from Simon Arlott.
1936 PP/09 New lookups functionality failed to compile on old gcc which rejects
1937 extern declarations in function scope.
1938 Patch from Oliver Fleischmann
1940 PP/10 Use sig_atomic_t for flags set from signal handlers.
1941 Check getgroups() return and improve debugging.
1942 Fixed developed for diagnosis in bug 927 (which turned out to be
1945 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1946 Patch from Mark Zealey.
1948 PP/12 Bugzilla 1056: Improved spamd server selection.
1949 Patch from Mark Zealey.
1951 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1952 Based on patch from Heiko Schlittermann.
1954 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1955 Patch from Uwe Doering, sign-off by Michael Haardt.
1957 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1958 variable declaration deep within a block. Bug and patch from
1961 PP/15 lookups-Makefile IRIX compatibility coercion.
1963 PP/16 Make DISABLE_DKIM build knob functional.
1965 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1966 Patch by Simon Arlott
1968 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1969 variable argument macros. Our copy now differs from upstream.
1975 TF/01 Failure to get a lock on a hints database can have serious
1976 consequences so log it to the panic log.
1978 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1979 controlled using the smtp_confirmation log selector.
1981 TF/03 Include the error message when we fail to unlink a spool file.
1983 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1984 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1985 for maintaining out-of-tree patches for some time.
1987 PP/01 Bugzilla 139: Documentation and portability issues.
1988 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1989 Handle per-OS dynamic-module compilation flags.
1991 PP/02 Let /dev/null have normal permissions.
1992 The 4.73 fixes were a little too stringent and complained about the
1993 permissions on /dev/null. Exempt it from some checks.
1994 Reported by Andreas M. Kirchwitz.
1996 PP/03 Report version information for many libraries, including
1997 Exim version information for dynamically loaded libraries. Created
1998 version.h, now support a version extension string for distributors
1999 who patch heavily. Dynamic module ABI change.
2001 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2002 privilege escalation vulnerability whereby the Exim run-time user
2003 can cause root to append content of the attacker's choosing to
2006 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2009 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2010 If dropping privileges for untrusted macros, we disabled normal logging
2011 on the basis that it would fail; for the Exim run-time user, this is not
2012 the case, and it resulted in successful deliveries going unlogged.
2013 Fixed. Reported by Andreas Metzler.
2019 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2020 only prepend for the Resent-* case. Fixes regression introduced in
2021 Exim 4.70 by NM/22 for Bugzilla 607.
2023 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2024 increasing numbers of administrators be bitten by this.
2026 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2028 PP/03 Bugzilla 994: added openssl_options main configuration option.
2030 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2032 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2034 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2036 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2037 without a peer certificate, leading to a segfault because of an
2038 assumption that peers always have certificates. Be a little more
2039 paranoid. Problem reported by Martin Tscholak.
2041 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2042 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2043 NB: ClamAV planning to remove STREAM in "middle of 2010".
2044 CL also introduces -bmalware, various -d+acl logging additions and
2045 more caution in buffer sizes.
2047 PP/09 Implemented reverse_ip expansion operator.
2049 PP/10 Bugzilla 937: provide a "debug" ACL control.
2051 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2053 PP/12 Bugzilla 973: Implement --version.
2055 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2057 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2059 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2061 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2062 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2063 ignore trailing whitespace.
2065 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2067 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2070 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2071 Notification from Dr Andrew Aitchison.
2073 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2074 ExtendedDetectionInfo response format.
2075 Notification from John Horne.
2077 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2080 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2081 XSL and documented dependency on system catalogs, with examples of how
2084 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2087 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2088 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2089 configuration file which is writeable by the Exim user or group.
2091 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2092 of configuration files to cover files specified with the -C option if
2093 they are going to be used with root privileges, not just the default
2096 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2097 option (effectively making it always true).
2099 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2100 files to be used while preserving root privileges.
2102 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2103 that rogue child processes cannot use them.
2105 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2106 run-time user, instead of root.
2108 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2109 Exim run-time user without dropping privileges.
2111 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2112 result string, instead of calling string_vformat() twice with the same
2115 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2116 for other users. Others should always drop root privileges if they use
2117 -C on the command line, even for a whitelisted configure file.
2119 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2121 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2127 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2128 $data_path, and $header_path variables; fixed documentation bugs and
2131 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2132 exipick to access non-standard spools, including the "frozen" queue
2135 NM/01 Bugzilla 965: Support mysql stored procedures.
2136 Patch from Alain Williams
2138 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2140 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2141 Patch from Andreas Metzler
2143 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2144 Patch from Kirill Miazine
2146 NM/05 Bugzilla 671: Added umask to procmail example.
2148 JJ/03 installed exipick 20100323.0, fixing doc bug
2150 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2151 directory. Notification and patch from Dan Rosenberg.
2153 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2155 TK/02 Improve log output when DKIM signing operation fails.
2157 MH/01 Treat the transport option dkim_domain as a colon separated
2158 list, not as a single string, and sign the message with each element,
2159 omitting multiple occurences of the same signer.
2161 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2162 Bugzilla 985, 986. Patch by Simon Arlott
2164 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2165 Patch by Simon Arlott
2167 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2168 MBX locking. Notification from Dan Rosenberg.
2174 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2176 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2178 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2180 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2182 NM/04 Bugzilla 915: Flag broken perl installation during build.
2188 TK/01 Added patch by Johannes Berg that expands the main option
2189 "spamd_address" if it starts with a dollar sign.
2191 TK/02 Write list of recipients to X-Envelope-Sender header when building
2192 the mbox-format spool file for content scanning (suggested by Jakob
2195 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2196 (http://www.dcc-servers.net/) support via dccifd. Activated by
2197 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2199 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2200 by Mark Daniel Reidel <mr@df.eu>.
2202 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2203 When building exim an external PCRE library is now needed -
2204 PCRE is a system library on the majority of modern systems.
2205 See entry on PCRE_LIBS in EDITME file.
2207 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2208 conversation. Added nologin parameter to request.
2209 Patch contributed by Kirill Miazine.
2211 TF/01 Do not log submission mode rewrites if they do not change the address.
2213 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2215 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2216 log files in place. Contributed by Roberto Lima.
2218 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2220 TF/03 Bugzilla 615: When checking the local_parts router precondition
2221 after a local_part_suffix or local_part_prefix option, Exim now
2222 does not use the address's named list lookup cache, since this
2223 contains cached lookups for the whole local part.
2225 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2226 Robert Millan. Documentation is in experimental-spec.txt.
2228 TF/04 Bugzilla 668: Fix parallel build (make -j).
2230 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2232 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2233 Patch provided by Jan Srzednicki.
2235 TF/05 Leading white space used to be stripped from $spam_report which
2236 wrecked the formatting. Now it is preserved.
2238 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2239 that they are available at delivery time.
2241 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2243 TF/08 TLS error reporting now respects the incoming_interface and
2244 incoming_port log selectors.
2246 TF/09 Produce a more useful error message if an SMTP transport's hosts
2247 setting expands to an empty string.
2249 NM/06 Bugzilla 744: EXPN did not work under TLS.
2250 Patch provided by Phil Pennock.
2252 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2253 Patch provided by Richard Godbee.
2255 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2256 acl_smtp_notquit, added index entry.
2258 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2259 Patch provided by Eugene Bujak.
2261 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2262 accept(). Patch provided by Maxim Dounin.
2264 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2265 Patch provided by Phil Pennock.
2267 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2269 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2270 Patch provided by Brad "anomie" Jorsch.
2272 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2273 Patch provided by Dean Brooks.
2275 TK/05 Add native DKIM support (does not depend on external libraries).
2277 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2278 Patch provided by Graeme Fowler.
2280 NM/16 Bugzilla 851: Documentation example syntax fix.
2282 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2284 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2287 NM/19 Bugzilla 745: TLS version reporting.
2288 Patch provided by Phil Pennock.
2290 NM/20 Bugzilla 167: bool: condition support.
2291 Patch provided by Phil Pennock.
2293 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2294 clients. Patch provided by Phil Pennock.
2296 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2297 Patch provided by Brad "anomie" Jorsch.
2299 NM/23 Bugzilla 687: Fix misparses in eximstats.
2300 Patch provided by Heiko Schlittermann.
2302 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2303 Patch provided by Heiko Schlittermann.
2305 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2306 plus update to original patch.
2308 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2310 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2311 Patch provided by David Brownlee.
2313 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2315 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2317 NM/30 Bugzilla 888: TLS documentation bugfixes.
2319 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2321 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2322 Unlike the original bugzilla I have changed all shell scripts in src tree.
2324 NM/33 Bugzilla 898: Transport filter timeout fix.
2325 Patch by Todd Rinaldo.
2327 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2328 Patch by Serge Demonchaux.
2330 NM/35 Bugzilla 39: Base64 decode bug fixes.
2331 Patch by Jakob Hirsch.
2333 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2335 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2337 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2339 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2345 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2346 ALT-N's libdkim that I have put here:
2347 http://duncanthrax.net/exim-experimental/
2349 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2350 were called 'true' and it seems that C99 defines that as a reserved
2351 keyword to be used with 'bool' variable types. That means you could
2352 not include C99-style headers which use bools without triggering
2353 build errors in sieve.c.
2355 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2356 as mailq or other aliases. Changed the --help handling significantly
2357 to do whats expected. exim_usage() emits usage/help information.
2359 SC/01 Added the -bylocaldomain option to eximstats.
2361 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2363 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2365 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2371 PH/01 Another patch from the Sieve maintainer.
2373 PH/02 When an IPv6 address is converted to a string for single-key lookup
2374 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2375 dots are used instead of colons so that keys in lsearch files need not
2376 contain colons. This was done some time before quoting was made available
2377 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2378 (notated using the quote facility) so as to distinguish them from IPv4
2379 keys. This meant that lookups for IP addresses in host lists did not work
2380 for iplsearch lookups.
2382 This has been fixed by arranging for IPv6 addresses to be expressed with
2383 colons if the lookup type is iplsearch. This is not incompatible, because
2384 previously such lookups could never work.
2386 The situation is now rather anomalous, since one *can* have colons in
2387 ordinary lsearch keys. However, making the change in all cases is
2388 incompatible and would probably break a number of configurations.
2390 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2393 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2394 conversion specification without a maximum field width, thereby enabling
2395 a rogue spamd server to cause a buffer overflow. While nobody in their
2396 right mind would setup Exim to query an untrusted spamd server, an
2397 attacker that gains access to a server running spamd could potentially
2398 exploit this vulnerability to run arbitrary code as the Exim user.
2400 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2401 $primary_hostname instead of what libspf2 thinks the hosts name is.
2403 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2404 a directory entry by the name of the lookup key. Previously, if a
2405 symlink pointed to a non-existing file or a file in a directory that
2406 Exim lacked permissions to read, a lookup for a key matching that
2407 symlink would fail. Now it is enough that a matching directory entry
2408 exists, symlink or not. (Bugzilla 503.)
2410 PH/03 The body_linecount and body_zerocount variables are now exported in the
2413 PH/04 Added the $dnslist_matched variable.
2415 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2416 This means they are set thereafter only if the connection becomes
2419 PH/06 Added the client_condition to authenticators so that some can be skipped
2420 by clients under certain conditions.
2422 PH/07 The error message for a badly-placed control=no_multiline_responses left
2423 "_responses" off the end of the name.
2425 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2427 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2428 (without spaces) instead of just copying the configuration text.
2430 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2432 PH/11 Added $max_received_linelength.
2434 PH/12 Added +ignore_defer and +include_defer to host lists.
2436 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2437 way in which PCRE > 7.0 is built.
2439 PH/14 Implemented queue_only_load_latch.
2441 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2442 MAIL command. The effect was to mangle the value on 64-bit systems.
2444 PH/16 Another patch from the Sieve maintainer.
2446 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2448 PH/18 If a system quota error occurred while trying to create the file for
2449 a maildir delivery, the message "Mailbox is full" was not appended to the
2450 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2451 only to a quota excession during the actual writing of the file.
2453 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2454 characters?) which causes problems in log lines. The DN values are now
2455 passed through string_printing() before being added to log lines.
2457 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2458 and InterBase are left for another time.)
2460 PH/21 Added message_body_newlines option.
2462 PH/22 Guard against possible overflow in moan_check_errorcopy().
2464 PH/23 POSIX allows open() to be a macro; guard against that.
2466 PH/24 If the recipient of an error message contained an @ in the local part
2467 (suitably quoted, of course), incorrect values were put in $domain and
2468 $local_part during the evaluation of errors_copy.
2474 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2475 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2478 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2479 issue a MAIL command.
2481 PH/02 In an ACL statement such as
2483 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2485 if a client was not listed at all, or was listed with a value other than
2486 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2487 the condition was not true (as it should be), so access was not denied.
2488 The bug was that the ! inversion was incorrectly passed on to the second
2489 item. This has been fixed.
2491 PH/03 Added additional dnslists conditions == and =& which are different from
2492 = and & when the dns lookup returns more than one IP address.
2494 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2495 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2497 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2498 FSYNC, which compiles an option called disable_fsync that allows for
2499 bypassing fsync(). The documentation is heavily laced with warnings.
2501 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2503 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2504 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2505 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2506 including adding "make clean"; (3) Added -fPIC when compiling the test
2507 dynamically loaded module, to get rid of a warning.
2509 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2510 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2511 = 0s. The bug is otherwise harmless.
2513 PH/07 There was a bug in the dovecot authenticator such that the value of
2514 $auth1 could be overwritten, and so not correctly preserved, after a
2515 successful authentication. This usually meant that the value preserved by
2516 the server_setid option was incorrect.
2518 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2520 PH/09 Installed PCRE release 7.0.
2522 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2523 run for batched SMTP input. It is now run at the start of every message
2524 in the batch. While fixing this I discovered that the process information
2525 (output by running exiwhat) was not always getting set for -bs and -bS
2526 input. This is fixed, and it now also says "batched" for BSMTP.
2528 PH/11 Added control=no_pipelining.
2530 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2531 patch, slightly modified), and move the expansion of helo_data till after
2532 the connection is made in the smtp transport (so it can use these
2535 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2537 PH/14 Added log_selector = +pid.
2539 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2541 PH/16 Add ${if forany and ${if forall.
2543 PH/17 Added dsn_from option to vary the From: line in DSNs.
2545 PH/18 Flush SMTP output before performing a callout, unless control =
2546 no_callout_flush is set.
2548 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2549 was true (the default) a successful delivery failed to delete the retry
2550 item, thus causing premature timeout of the address. The bug is now
2553 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2555 PH/21 Long custom messages for fakedefer and fakereject are now split up
2556 into multiline responses in the same way that messages for "deny" and
2557 other ACL rejections are.
2559 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2560 with slight modification.
2562 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2563 draft, changing the syntax and factoring some duplicate code".
2565 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2566 for deliveries of the second and subsequent messages over the same SMTP
2569 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2570 ${reduce, with only minor "tidies".
2572 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2574 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2575 expansion side effects.
2577 PH/27 When a message times out after an over-quota error from an Exim-imposed
2578 quota, the bounce message says "mailbox is full". This message was not
2579 being given when it was a system quota that was exceeded. It now should
2582 MH/03 Made $recipients available in local_scan(). local_scan() already has
2583 better access to the recipient list through recipients_list[], but
2584 $recipients can be useful in postmaster-provided expansion strings.
2586 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2587 in the case of a MAIL command with additional options following the
2588 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2589 were accidentally chopped off.
2591 PH/29 SMTP synchronization checks are implemented when a command is read -
2592 there is a check that no more input is waiting when there shouldn't be
2593 any. However, for some commands, a delay in an ACL can mean that it is
2594 some time before the response is written. In this time, more input might
2595 arrive, invalidly. So now there are extra checks after an ACL has run for
2596 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2597 pipelining has not been advertised.
2599 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2601 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2602 returned in the SMTP response when smtp_return_error_details was false.
2603 This has been fixed.
2605 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2606 instead of the C I/O that was originally supplied, because problems were
2607 reported on Solaris.
2609 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2610 Exim which did not show up earlier: it was assuming that a call to
2611 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2612 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2613 was a macro that became an assignment, so it seemed to work. This has
2614 changed to a proper function call with a void return, hence the compile
2615 error. Exim's code has been fixed.
2617 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2620 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2622 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2624 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2625 because it turns out to be too big or there is a timeout) there is no
2626 "Completed" line in the log. When some messages of this type were
2627 selected by exigrep, they were listed as "not completed". Others were
2628 picked up by some special patterns. I have improved the selection
2629 criteria to be more general.
2631 PH/38 The host_find_failed option in the manualroute router can now be set
2632 to "ignore", to completely ignore a host whose IP address cannot be
2633 found. If all hosts are ignored, the behaviour is controlled by the new
2634 host_all_ignored option.
2636 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2637 homing or because of multiple MX records with /mx) generated more than
2638 one IP address, and the following item turned out to be the local host,
2639 all the secondary addresses of the first item were incorrectly removed
2640 from the list, along with the local host and any following hosts (which
2641 is what is supposed to happen).
2643 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2644 whoever called Exim into the -H file. In the case of the daemon it was
2645 behaving confusingly. When first started, it used values for whoever
2646 started the daemon, but after a SIGHUP it used the Exim user (because it
2647 calls itself on a restart). I have changed the code so that it now always
2650 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2651 message are rejected with the same error (e.g. no authentication or bad
2652 sender address), and a DATA command is nevertheless sent (as can happen
2653 with PIPELINING or a stupid MUA), the error message that was given to the
2654 RCPT commands is included in the rejection of the DATA command. This is
2655 intended to be helpful for MUAs that show only the final error to their
2658 PH/42 Another patch from the Sieve maintainer.
2660 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2661 Eximstats - Fixed some broken HTML links and added missing column headers
2663 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2664 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2666 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2668 PH/43 Yet another patch from the Sieve maintainer.
2670 PH/44 I found a way to check for a TCP/IP connection going away before sending
2671 the response to the final '.' that terminates a message, but only in the
2672 case where the client has not sent further data following the '.'
2673 (unfortunately, this is allowed). However, in many cases there won't be
2674 any further data because there won't be any more messages to send. A call
2675 to select() can be used: if it shows that the input is "ready", there is
2676 either input waiting, or the socket has been closed. An attempt to read
2677 the next input character can distinguish the two cases. Previously, Exim
2678 would have sent an OK response which the client would never have see.
2679 This could lead to message repetition. This fix should cure that, at
2680 least in a lot of common cases.
2682 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2683 advertised in response to EHLO.
2689 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2690 fixed by 4.65/MH/01 (is this a record?) are fixed:
2692 (i) An empty string was always treated as zero by the numeric comparison
2693 operators. This behaviour has been restored.
2695 (ii) It is documented that the numeric comparison operators always treat
2696 their arguments as decimal numbers. This was broken in that numbers
2697 starting with 0 were being interpreted as octal.
2699 While fixing these problems I realized that there was another issue that
2700 hadn't been noticed. Values of message_size_limit (both the global option
2701 and the transport option) were treated as octal if they started with 0.
2702 The documentation was vague. These values are now always treated as
2703 decimal, and I will make that clear in the documentation.
2709 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2710 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2713 MH/01 Don't check that the operands of numeric comparison operators are
2714 integers when their expansion is in "skipping" mode (fixes bug
2715 introduced by 4.64-PH/07).
2717 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2718 child addresses, Exim now panics and dies. Previously, because the count
2719 is held in a short int, deliveries were likely to be lost. As such a
2720 large number of recipients for a single message is ridiculous
2721 (performance will be very, very poor), I have chosen to impose a limit
2722 rather than extend the field.
2728 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2729 leftover -K file (the existence of which was triggered by #402).
2730 While we were at it, introduced process PID as part of the -K
2731 filename. This should rule out race conditions when creating
2734 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2735 processing considerably. Previous code took too long for large mails,
2736 triggering a timeout which in turn triggers #401.
2738 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2739 in the DK code in transports.c. sendfile() is not really portable,
2740 hence the _LINUX specificness.
2742 TF/01 In the add_headers option to the mail command in an Exim filter,
2743 there was a bug that Exim would claim a syntax error in any
2744 header after the first one which had an odd number of characters
2747 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2748 callout verification, Exim cached a "reject" for the entire domain. This
2749 is correct for most verifications, but it is not correct for a recipient
2750 verification with use_sender or use_postmaster set, because in that case
2751 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2752 case of MAIL FROM:<> rejection from other early rejections (e.g.
2753 rejection of HELO). When verifying a recipient using a non-null MAIL
2754 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2755 Whatever the result of the callout, the value of the domain cache is
2756 left unchanged (for any other kind of callout, getting as far as trying
2757 RCPT means that the domain itself is ok).
2759 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2762 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2763 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2764 session, the daemon ignores ECONNECT errors and logs others; it now
2765 ignores EPIPE as well.
2767 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2768 (quoted-printable decoding).
2770 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2771 later the small subsequent patch to fix an introduced bug.
2773 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2775 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2777 PH/08 An error is now given if message_size_limit is specified negative.
2779 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2780 to be given (somewhat) arbitrary names.
2782 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2785 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2786 miscellaneous code fixes
2788 PH/10 Added the log_reject_target ACL modifier to specify where to log
2791 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2792 hostname. This is wrong, because it relates to the incoming message (and
2793 probably the interface on which it is arriving) and not to the outgoing
2794 callout (which could be using a different interface). This has been
2795 changed to use the value of the helo_data option from the smtp transport
2796 instead - this is what is used when a message is actually being sent. If
2797 there is no remote transport (possible with a router that sets up host
2798 addresses), $smtp_active_hostname is used.
2800 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2801 tweaks were necessary in order to get it to work (see also 21 below):
2802 (a) The code assumed that strncpy() returns a negative number on buffer
2803 overflow, which isn't the case. Replaced with Exim's string_format()
2805 (b) There were several signed/unsigned issues. I just did the minimum
2806 hacking in of casts. There is scope for a larger refactoring.
2807 (c) The code used strcasecmp() which is not a standard C function.
2808 Replaced with Exim's strcmpic() function.
2809 (d) The code set only $1; it now sets $auth1 as well.
2810 (e) A simple test gave the error "authentication client didn't specify
2811 service in request". It would seem that Dovecot has changed its
2812 interface. Fortunately there's a specification; I followed it and
2813 changed what the client sends and it appears to be working now.
2815 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2818 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2819 address A is aliased to B and C, where B exists and C does not. Without
2820 -v the output is "A verified" because verification stops after a
2821 successful redirection if more than one address is generated. However,
2822 with -v the child addresses are also verified. Exim was outputting "A
2823 failed to verify" and then showing the successful verification for C,
2824 with its parentage. It now outputs "B failed to verify", showing B's
2825 parentage before showing the successful verification of C.
2827 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2828 look up a TXT record in a specific list after matching in a combined
2831 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2832 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2833 they consult the DNS. I had assumed they would set it the way they
2834 wanted; and indeed my experiments on Linux seem to show that in some
2835 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2836 To be on the safe side, however, I have now made the interface to
2837 host_find_byname() similar to host_find_bydns(), with an argument
2838 containing the DNS resolver options. The host_find_byname() function now
2839 sets these options at its start, just as host_find_bydns() does. The smtp
2840 transport options dns_qualify_single and dns_search_parents are passed to
2841 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2842 of host_find_byname() use the default settings of RES_DEFNAMES
2843 (qualify_single) but not RES_DNSRCH (search_parents).
2845 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2846 spool_read_header() do less string testing, by means of a preliminary
2847 switch on the second character of optional "-foo" lines. (This is
2848 overdue, caused by the large number of possibilities that now exist.
2849 Originally there were few.) While I was there, I also converted the
2850 str(n)cmp tests so they don't re-test the leading "-" and the first
2851 character, in the hope this might squeeze out yet more improvement.
2853 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2854 flag allowing group syntax was set by the header_syntax check but not
2855 turned off, possible causing trouble later; (2) The flag was not being
2856 set at all for the header_verify test, causing "group"-style headers to
2857 be rejected. I have now set it in this case, and also caused header_
2858 verify to ignore an empty address taken from a group. While doing this, I
2859 came across some other cases where the code for allowing group syntax
2860 while scanning a header line wasn't quite right (mostly, not resetting
2861 the flag correctly in the right place). These bugs could have caused
2862 trouble for malformed header lines. I hope it is now all correct.
2864 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2865 with the "reply" argument non-NULL. The code, however (which originally
2866 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2867 but it didn't always do it. This confused somebody who was copying the
2868 code for some other use. I have removed all the tests.
2870 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2871 feature that was used to support insecure browsers during the U.S. crypto
2872 embargo. It requires special client support, and Exim is probably the
2873 only MTA that supported it -- and would never use it because real RSA is
2874 always available. This code has been removed, because it had the bad
2875 effect of slowing Exim down by computing (never used) parameters for the
2876 RSA_EXPORT functionality.
2878 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2879 authenticator to fail if there's a tab character in the incoming data
2880 (there should never be unless someone is messing about, as it's supposed
2881 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2882 if the connection is using TLS or if the remote IP is the same as the
2883 local IP, and the "valid-client-cert option" if a client certificate has
2886 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2887 authenticators. This can be used for authorization after authentication
2888 succeeds. (In the case of plaintext, it servers for both authentication
2891 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2892 if any retry times were supplied.
2894 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2895 connection, where there is no client IP address to check. In this
2896 situation, the verify now always succeeds.
2898 PH/25 Applied John Jetmore's -Mset patch.
2900 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2902 PH/27 In a string expansion for a processed (not raw) header when multiple
2903 headers of the same name were present, leading whitespace was being
2904 removed from all of them, but trailing whitespace was being removed only
2905 from the last one. Now trailing whitespace is removed from each header
2906 before concatenation. Completely empty headers in a concatenation (as
2907 before) are ignored.
2909 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2910 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2912 PH/29 [Removed. This was a change that I later backed out, and forgot to
2913 correct the ChangeLog entry (that I had efficiently created) before
2914 committing the later change.]
2916 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2917 address errors (4xx response to RCPT) over the same connection as other
2918 messages routed to the same hosts. Such deliveries are always "forced",
2919 so retry times are not inspected. This resulted in far too many retries
2920 for the affected addresses. The effect occurred only when there were more
2921 hosts than the hosts_max_try setting in the smtp transport when it had
2922 the 4xx errors. Those hosts that it had tried were not added to the list
2923 of hosts for which the message was waiting, so if all were tried, there
2924 was no problem. Two fixes have been applied:
2926 (i) If there are any address or message errors in an SMTP delivery, none
2927 of the hosts (tried or untried) are now added to the list of hosts
2928 for which the message is waiting, so the message should not be a
2929 candidate for sending over the same connection that was used for a
2930 successful delivery of some other message. This seems entirely
2931 reasonable: after all the message is NOT "waiting for some host".
2932 This is so "obvious" that I'm not sure why it wasn't done
2933 previously. Hope I haven't missed anything, but it can't do any
2934 harm, as the worst effect is to miss an optimization.
2936 (ii) If, despite (i), such a delivery is accidentally attempted, the
2937 routing retry time is respected, so at least it doesn't keep
2938 hammering the server.
2940 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2941 in ${readsocket because some servers need this prod.
2943 PH/32 Added some extra debug output when updating a wait-xxx database.
2945 PH/33 The hint "could be header name not terminated by colon", which has been
2946 given for certain expansion errors for a long time, was not being given
2947 for the ${if def:h_colon_omitted{... case.
2949 PH/34 The spec says: "With one important exception, whenever a domain list is
2950 being scanned, $domain contains the subject domain." There was at least
2951 one case where this was not true.
2953 PH/35 The error "getsockname() failed: connection reset by peer" was being
2954 written to the panic log as well as the main log, but it isn't really
2955 panic-worthy as it just means the connection died rather early on. I have
2956 removed the panic log writing for the ECONNRESET error when getsockname()
2959 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2960 runs only) independently of the message's sender address. This meant
2961 that, if the 4xx error was in fact related to the sender, a different
2962 message to the same recipient with a different sender could confuse
2963 things. In particular, this can happen when sending to a greylisting
2964 server, but other circumstances could also provoke similar problems.
2965 I have changed the default so that the retry time for these errors is now
2966 based a combination of the sender and recipient addresses. This change
2967 can be overridden by setting address_retry_include_sender=false in the
2970 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2971 remote server are returned as part of bounce messages. This was not
2972 happening for LMTP over a pipe (the lmtp transport), but now it is the
2973 same for both kinds of LMTP.
2975 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2976 in header lines that were in fact CNAMEs. This is no longer the case.
2978 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2979 and queue runs started by the daemon processed all messages. This has
2980 been fixed so that -R and -S can now usefully be given with -q<time>.
2982 PH/40 Import PCRE release 6.7 (fixes some bugs).
2984 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2986 PH/42 Give an error if -q is specified more than once.
2988 PH/43 Renamed the variables $interface_address and $interface_port as
2989 $received_ip_address and $received_port, to make it clear that these
2990 values apply to message reception, and not to the outgoing interface when
2991 a message is delivered. (The old names remain recognized, of course.)
2993 PH/44 There was no timeout on the connect() call when using a Unix domain
2994 socket in the ${readsocket expansion. There now is.
2996 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2997 be meaningful with "accept".
2999 SC/01 Eximstats V1.43
3000 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3002 SC/02 Eximstats V1.44
3003 Use a glob alias rather than an array ref in the generated
3004 parser. This improves both readability and performance.
3006 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3007 Collect SpamAssassin and rejection statistics.
3008 Don't display local sender or destination tables unless
3009 there is data to show.
3010 Added average volumes into the top table text output.
3012 SC/04 Eximstats V1.46
3013 Collect data on the number of addresses (recipients)
3014 as well as the number of messages.
3016 SC/05 Eximstats V1.47
3017 Added 'Message too big' to the list of mail rejection
3018 reasons (thanks to Marco Gaiarin).
3020 SC/06 Eximstats V1.48
3021 Mainlog lines which have GMT offsets and are too short to
3022 have a flag are now skipped.
3024 SC/07 Eximstats V1.49 (Alain Williams)
3025 Added the -emptyok flag.
3027 SC/08 Eximstats V1.50
3028 Fixes for obtaining the IP address from reject messages.
3030 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3031 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3032 whitespace changes from 4.64-PH/27
3034 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3037 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3038 are found, allow negative numbers in numeric criteria)
3040 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3042 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3043 to match changes made in 4.64-PH/43
3045 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3047 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3048 "domain" is actually an IP address, and force a failure. This locks out
3049 those revolvers/nameservers that support "A-for-A" lookups, in
3050 contravention of the specifications.
3052 PH/48 When a host name was looked up from an IP address, and the subsequent
3053 forward lookup of the name timed out, the host name was left in
3054 $sender_host_name, contrary to the specification.
3056 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3057 restricted to single-key lookups, Exim was not diagnosing an error if
3058 * or *@ was used with a query-style lookup.
3060 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3062 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3063 long ago, but noone interested enough thought of it. Let's just say that
3064 the "1.1" means that there are some new functions that weren't there at
3065 some point in the past.
3067 PH/51 Error processing for expansion failure of helo_data from an smtp
3068 transport during callout processing was broken.
3070 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3071 tested/used via the -bh/-bhc/-bs options.
3073 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3074 bug, fixed in subsequent PCRE releases).
3076 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3077 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3079 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3085 SC/01 Use a glob alias rather than an array ref in eximstats generated
3086 parser. This improves both readability and performance.
3088 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3089 Don't display local sender or destination tables in eximstats unless
3090 there is data to show.
3091 Added average volumes into the eximstats top table text output.
3093 SC/03 Collect data on the number of addresses (recipients) as well
3094 as the number of messages in eximstats.
3096 TF/01 Correct an error in the documentation for the redirect router. Exim
3097 does (usually) call initgroups() when daemonizing.
3099 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3100 with consistent privilege compared to when running as a daemon.
3102 TF/03 Note in the spec that $authenticated_id is not set for local
3103 submissions from trusted users.
3105 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3106 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3108 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3109 by adding some example configuration directives to the default
3110 configuration file. A little bit of work is required to uncomment the
3111 directives and define how usernames and passwords are checked, but
3112 there is now a framework to start from.
3114 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3115 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3116 without this. I don't know how relevant this is to other LDAP libraries.
3118 PH/02 Add the verb name to the "unknown ACL verb" error.
3120 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3122 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3124 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3125 directory not expanded when it should be if an expanded home directory
3126 was set for the address (which is overridden by the transport).
3128 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3131 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3132 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3133 because it is too late at that time, and has no effect.
3135 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3136 security issue with \' (bugzilla #107). I could not use the
3137 PQescapeStringConn() function, because it needs a PGconn value as one of
3140 PH/08 When testing addresses using -bt, indicate those final addresses that
3141 are duplicates that would not cause an additional delivery. At least one
3142 person was confused, thinking that -bt output corresponded to deliveries.
3143 (Suppressing duplicates isn't a good idea as you lose the information
3144 about possibly different redirections that led to the duplicates.)
3146 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3147 systems where poll() doesn't work, in particular OS X.
3149 PH/10 Added more information to debugging output for retry time not reached.
3151 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3152 operations in malware.c.
3154 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3157 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3158 syslog_duplication set false, log lines that would normally be written
3159 both the the main log and to the reject log were not written to syslog at
3162 PH/14 In the default configuration, change the use of "message" in ACL warn
3163 statements to "add_header".
3165 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3166 not followed by a command (e.g. "seen endif").
3168 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3169 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3172 PH/17 Added extra conditions to the default value of delay_warning_condition
3176 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3177 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3178 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3181 The Auto-Submitted: and various List- headers are standardised, whereas I
3182 don't think Precedence: ever was.
3184 PH/18 Refactored debugging code in route_finduser() to show more information,
3185 in particular, the error code if getpwnam() issues one.
3187 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3188 This is apparently needed in addition to the PH/07 change above to avoid
3189 any possible encoding problems.
3191 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3192 but not after initializing Perl.
3194 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3195 output them only if debugging. By default they are written stderr,
3196 apparently, which is not desirable.
3198 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3201 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3204 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3206 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3207 authenticated or an ident call has been made. Suppress the default
3208 values for $authenticated_id and $authenticated_sender (but permit -oMai
3209 and -oMas) when testing with -bh.
3211 PH/24 Re-jigged the order of the tests in the default configuration so that the
3212 tests for valid domains and recipients precede the DNS black list and CSA
3213 tests, on the grounds that those ones are more expensive.
3215 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3216 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3217 This bug exists in every version of Exim that I still have, right back to
3220 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3221 However, an attempt to turn on case-sensitivity in a regex key by
3222 including (?-i) didn't work because the subject string was already
3223 lowercased, and the effects were non-intuitive. It turns out that a
3224 one-line patch can be used to allow (?-i) to work as expected.
3230 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3231 other effects) broke the use of negated acl sub-conditions.
3233 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3236 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3237 "Deny" causes Exim to reject the incoming connection with a 554 error.
3238 Unfortunately, if there is a major crisis, such as a disk failure,
3239 tcp-wrappers gives "deny", whereas what one would like would be some
3240 kind of temporary error. A kludge has been added to help with this.
3241 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3242 554 error is used if errno is still zero or contains ENOENT (which occurs
3243 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3246 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3248 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3249 errors. Otherwise a message that provokes a temporary error (when other
3250 messages do not) can cause a whole host to time out.
3252 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3253 addresses were routed directly to files or pipes from a redirect router.
3254 File deliveries just didn't batch; pipe deliveries might have suffered
3257 PH/06 A failure to get a lock for a hints database would erroneously always say
3258 "Failed to get write lock", even when it was really a read lock.
3260 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3261 of 0600. This has been changed to use the value of the lockfile_mode
3262 option (which defaults to 0600).
3264 PH/08 Applied small patch from the Sieve maintainer.
3266 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3267 folder from quota calculations, a direct delivery into this folder messed
3268 up the contents of the maildirsize file. This was because the regex was
3269 used only to exclude .Trash (or whatever) when the size of the mailbox
3270 was calculated. There was no check that a delivery was happening into an
3271 excluded directory. This bug has been fixed by ignoring all quota
3272 processing for deliveries into excluded directories.
3274 PH/10 Added the maildirfolder_create_regex option to appendfile.
3280 PH/01 The code for finding all the local interface addresses on a FreeBSD
3281 system running IPv6 was broken. This may well have applied to all BSD
3282 systems, as well as to others that have similar system calls. The broken
3283 code found IPv4 interfaces correctly, but gave incorrect values for the
3284 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3285 that it would not match correctly against @[] and not recognize the IPv6
3288 PH/02 The ipliteral router was not recognizing addresses of the form user@
3289 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3291 PH/03 Added disable_ipv6.
3293 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3294 decoded form, because it is most often used to construct To: headers
3295 lines in autoreplies, and the decoded form may well be syntactically
3296 invalid. However, $reply_address has leading white space removed, and all
3297 newlines turned into spaces so that the autoreply transport does not
3300 PH/05 If group was specified without a user on a router, and no group or user
3301 was specified on a transport, the group from the router was ignored.
3303 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3304 for visible compile-time settings that can be used to change these
3305 numbers, for those that want even more. Backwards compatibility with old
3306 spool files has been maintained. However, going back to a previous Exim
3307 release will lost any variables that are in spool files.
3309 PH/07 Two small changes when running in the test harness: increase delay when
3310 passing a TCP/IP connection to a new process, in case the original
3311 process has to generate a bounce, and remove special handling of
3312 127.0.0.2 (sic), which is no longer necessary.
3314 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3315 be the same on different OS.
3317 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3320 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3321 whether --show-vars was specified or not
3323 JJ/02 exipick: Added support for new ACL variable spool format introduced
3326 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3327 syntactically invalid From: or Reply-to: line, and a filter used this to
3328 generate an autoreply, and therefore failed to obtain an address for the
3329 autoreply, Exim could try to deliver to a non-existent relative file
3330 name, causing unrelated and misleading errors. What now happens is that
3331 it logs this as a hard delivery error, but does not attempt to create a
3334 PH/11 The exinext utility has a -C option for testing purposes, but although
3335 the given file was scanned by exinext itself; it wasn't being passed on
3336 when Exim was called.
3338 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3339 an end-of-file indication when reading a command response.
3341 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3342 compiled. In many other places in Exim, IPv6 addresses are always
3343 recognized, so I have changed this. It also means that IPv4 domain
3344 literals of the form [IPV4:n.n.n.n] are now always recognized.
3346 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3347 used if the router is not running as root, for example, when verifying at
3348 ACL time, or when using -bh. The debugging output from this situation was
3349 non-existent - all you got was a failure to exec. I have made two
3352 (a) Failures to set uid/gid, the current directory, or a process leader
3353 in a subprocess such as that created by queryprogram now generate
3354 suitable debugging output when -d is set.
3356 (b) The queryprogram router detects when it is not running as root,
3357 outputs suitable debugging information if -d is set, and then runs
3358 the subprocess without attempting to change uid/gid.
3360 PH/15 Minor change to Makefile for building test_host (undocumented testing
3363 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3364 additional section of a DNS packet that returns MX or SRV records.
3365 Instead, it always explicitly searches for A/AAAA records. This avoids
3366 major problems that occur when a DNS server includes only records of one
3367 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3368 fixed another bug: if SRV records were looked up and the corresponding
3369 address records were *not* found in the additional section, the port
3370 values from the SRV records were lost.
3372 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3373 using the correct key (the original address) when searching the retry
3374 rules in order to find which one to use for generating the retry hint.
3376 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3377 adding the default one. Similarly, if it contains a Reply-To: header, the
3378 errors_reply_to option, if set, is not used.
3380 PH/19 When calculating a retry time, Exim used to measure the "time since
3381 failure" by looking at the "first failed" field in the retry record. Now
3382 it does not use this if it is later than than the arrival time of the
3383 message. Instead it uses the arrival time. This makes for better
3384 behaviour in cases where some deliveries succeed, thus re-setting the
3385 "first failed" field. An example is a quota failure for a huge message
3386 when small messages continue to be delivered. Without this change, the
3387 "time since failure" will always be short, possible causing more frequent
3388 delivery attempts for the huge message than are intended.
3389 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3391 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3392 $1, $2, $3) because the numerical variables can be reset during some
3393 expansion items (e.g. "match"), thereby losing the authentication data.
3395 PH/21 Make -bV show the size of off_t variables so that the test suite can
3396 decide whether to run tests for quotas > 2G.
3398 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3399 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3400 filecount value is greater than 2G or if a quota value is greater than 2G
3401 on a system where the size of off_t is not greater than 4, a panic error
3404 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3405 never match. The debug and -bh output now contains an explicit error
3406 message indicating a malformed IPv4 address or mask.
3408 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3409 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3410 PH/23 above applies.
3412 PH/25 Do not write to syslog when running in the test harness. The only
3413 occasion when this arises is a failure to open the main or panic logs
3414 (for which there is an explicit test).
3416 PH/26 Added the /no_tell option to "control=freeze".
3418 PH/27 If a host name lookup failed very early in a connection, for example, if
3419 the IP address matched host_lookup and the reverse lookup yielded a name
3420 that did not have a forward lookup, an error message of the form "no IP
3421 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3422 could be logged. Now it outputs the IP address instead of "NULL".
3424 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3425 allows the sender and the authenticated sender to be set when
3426 submitting a message from within Exim. Since child_open_exim() is
3427 documented for local_scan(), the new function should be too.
3429 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3430 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3431 results in an empty string is now treated as unset.
3433 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3435 PH/31 Added sender_verify_fail logging option.
3437 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3438 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3439 tidied the source and removed it altogether.
3441 PH/33 When a queue run was abandoned because the load average was too high, a
3442 log line was always written; now it is written only if the queue_run log
3443 selector is set. In addition, the log line for abandonment now contains
3444 information about the queue run such as the pid. This is always present
3445 in "start" and "stop" lines but was omitted from the "abandon" line.
3447 PH/34 Omit spaces between a header name and the colon in the error message that
3448 is given when verify = headers_syntax fails (if there are lots of them,
3449 the message gets confusing).
3451 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3452 names, as there are now some PTR records that contain slashes. This check
3453 is only to protect against broken name servers that fall over on strange
3454 characters, so the fact that it applies to all lookups doesn't matter.
3456 PH/36 Now that the new test suite is complete, we can remove some of the
3457 special code in Exim that was needed for the old test suite. For example,
3458 sorting DNS records because real resolvers return them in an arbitrary
3459 order. The new test suite's fake resolver always returns records in the
3462 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3463 bounces) except when queue_only is set, to avoid logging races between
3464 the different processes.
3466 PH/38 Panic-die if .include specifies a non-absolute path.
3468 PH/39 A tweak to the "H" retry rule from its user.
3470 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3471 a label. They prevented compilation on older perls.
3473 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3474 a warning to be raised on newish perls.
3476 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3477 on queue. Changes to match documented behaviour of showing count of
3478 messages matching specified criteria.
3480 PH/40 Changed the default ident timeout from 30s to 5s.
3482 PH/41 Added support for the use of login_cap features, on those BSD systems
3483 that have them, for controlling the resources used by pipe deliveries.
3485 PH/42 The content-scanning code uses fopen() to create files in which to put
3486 message data. Previously it was not paying any attention to the mode of
3487 the files. Exim runs with umask(0) because the rest of the code creates
3488 files with open(), and sets the required mode explicitly. Thus, these
3489 files were ending up world-writeable. This was not a big issue, because,
3490 being within the spool directory, they were not world-accessible. I have
3491 created a function called modefopen, which takes an additional mode
3492 argument. It sets umask(777), creates the file, chmods it to the required
3493 mode, then resets the umask. All the relevant calls to fopen() in the
3494 content scanning code have been changed to use this function.
3496 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3497 to 24 hours. This avoids potential overflow problems when processing G
3498 and H retry rules. I suspect nobody ever tinkers with this value.
3500 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3502 PH/45 When the plaintext authenticator is running as a client, the server's
3503 challenges are checked to ensure they are valid base64 strings. By
3504 default, the authentication attempt is cancelled if an invalid string is
3505 received. Setting client_ignore_invalid_base64 true ignores these errors.
3506 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3507 they are received. Thus, the responses can be made to depend on the
3508 challenges. If an invalid string is ignored, an empty string is placed in
3511 PH/46 Messages that are created by the autoreply transport now contains a
3512 References: header, in accordance with RFCs 2822 and 3834.
3514 PH/47 Added authenticated_sender_force to the smtp transport.
3516 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3518 PH/49 Installed latest patch from the Sieve maintainer.
3520 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3521 was also set, the appendfile transport was unnecessarily scanning a
3522 directory of message files (e.g. for maildir delivery) to find the count
3523 of files (along with the size), even though it did not need this
3524 information. It now does the scan only if it needs to find either the
3525 size of the count of files.
3527 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3529 PH/52 Two bugs concerned with error handling when the smtp transport is
3532 (i) Exim was not creating retry information for temporary errors given
3533 for individual recipients after the DATA command when the smtp transport
3534 was used in LMTP mode. This meant that they could be retried too
3535 frequently, and not timed out correctly.
3537 (ii) Exim was setting the flag that allows error details to be returned
3538 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3539 recipients that were returned after the DATA command.
3541 PH/53 This is related to PH/52, but is more general: for any failing address,
3542 when detailed error information was permitted to be returned to the
3543 sender, but the error was temporary, then after the final timeout, only
3544 "retry timeout exceeded" was returned. Now it returns the full error as
3545 well as "retry timeout exceeded".
3547 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3548 do this, and (what is worse) MTAs that accept it.
3550 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3551 will now be deprecated.
3553 PH/56 New os.c-cygwin from the Cygwin maintainer.
3555 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3556 formats (previously only available in exim formats via -bpr, -bpru,
3557 and -bpra. Now also available in native and exiqgrep formats)
3559 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3560 with very large, slow to parse queues
3562 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3564 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3566 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3567 responses to authentication challenges, though it was showing the
3568 challenges; (ii) I've removed the CR characters from the debug output for
3571 PH/58 Allow for the insertion of a newline as well as a space when a string
3572 is turned into more than one encoded-word during RFC 2047 encoding. The
3573 Sieve code now uses this.
3575 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3576 data_4xx, lost_connection, tls_required.
3578 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3579 message was being sent as an SMTP response.
3581 PH/61 Add -l and -k options to exicyclog.
3583 PH/62 When verifying, if an address was redirected to one new address, so that
3584 verification continued, and the new address failed or deferred after
3585 having set something in $address_data, the value of $address_data was not
3586 passed back to the ACL. This was different to the case when no
3587 redirection occurred. The value is now passed back in both cases.
3589 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3590 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3591 use login_cap.h, so on its own it isn't the distinguishing feature. The
3592 new name refers directly to the setclassresources() function.
3594 PH/65 Added configuration files for NetBSD3.
3596 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3598 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3599 is preferred over IPv4.
3601 PH/68 The bounce_return_message and bounce_return_body options were not being
3602 honoured for bounces generated during the reception of non-SMTP messages.
3603 In particular, this applied to messages rejected by the ACL. This bug has
3604 been fixed. However, if bounce_return_message is true and bounce_return_
3605 body is false, the headers that are returned for a non-SMTP message
3606 include only those that have been read before the error was detected.
3607 (In the case of an ACL rejection, they have all been read.)
3609 PH/69 The HTML version of the specification is now built in a directory called
3610 spec_html instead of spec.html, because the latter looks like a path with
3611 a MIME-type, and this confuses some software.
3613 PH/70 Catch two compiler warnings in sieve.c.
3615 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3616 function verify_get_ident() calls ip_connect() to connect a socket, but
3617 if the "connect()" function timed out, ip_connect() used to close the
3618 socket. However, verify_get_ident() also closes the socket later, and in
3619 between Exim writes to the log, which may get opened at this point. When
3620 the socket was closed in ip_connect(), the log could get the same file
3621 descriptor number as the socket. This naturally causes chaos. The fix is
3622 not to close the socket in ip_connect(); the socket should be closed by
3623 the function that creates it. There was only one place in the code where
3624 this was missing, in the iplookup router, which I don't think anybody now
3625 uses, but I've fixed it anyway.
3627 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3628 well as to direct DNS lookups. Otherwise the handling of names in host
3629 lists is inconsistent and therefore confusing.
3635 PH/01 Two changes to the default runtime configuration:
3637 (1) Move the checks for relay_from_hosts and authenticated clients from
3638 after to before the (commented out) DNS black list checks.
3640 (2) Add control=submission to the relay_from_hosts and authenticated
3641 clients checks, on the grounds that messages accepted by these
3642 statements are most likely to be submissions.
3644 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3646 (1) Generate an error if the third argument for the ${prvs expansion is
3649 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3652 (3) Reset the variables that are obtained from the first argument of
3653 ${prvscheck and used in the second argument before leaving the code,
3654 because their memory is reclaimed, so using them afterwards may do
3657 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3658 one (it's much easier than Tom thought :-).
3660 (5) Because of (4), we can now allow for the use of $prvscheck_result
3661 inside the third argument.
3663 PH/03 For some reason, the default setting of PATH when running a command from
3664 a pipe transport was just "/usr/bin". I have changed it to
3667 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3668 anything to be listed in the output from -bV.
3670 PH/05 When a filter generated an autoreply, the entire To: header line was
3671 quoted in the delivery log line, like this:
3673 => >A.N.Other <ano@some.domain> <original@ddress> ...
3675 This has been changed so that it extracts the operative address. There
3676 may be more than one such address. If so, they are comma-separated, like
3679 => >ano@some.domain,ona@other.domain <original@ddress> ...
3681 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3682 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3683 not being looked up in the rDNS to get a host name, Exim was showing the
3684 IP address twice in Received: lines, even though the IP addresses were
3685 identical. For example:
3687 Received: from [1.2.3.4] (helo=[1.2.3.4])
3689 However, if the real host name was known, it was omitting the HELO data
3690 if it matched the actual IP address. This has been tidied up so that it
3691 doesn't show the same IP address twice.
3693 PH/07 When both +timestamp and +memory debugging was on, the value given by
3694 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3695 called by the debug printing, thereby overwriting the timestamp buffer.
3696 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3698 PH/08 When the original message was included in an autoreply transport, it
3699 always said "this is a copy of the message, including all the headers",
3700 even if body_only or headers_only was set. It now gives an appropriate
3703 PH/09 Applied a patch from the Sieve maintainer which:
3705 o fixes some comments
3706 o adds the (disabled) notify extension core
3707 o adds some debug output for the result of if/elsif tests
3708 o points to the current vacation draft in the documentation
3709 and documents the missing references header update
3713 o fixes a bug in processing the envelope test (when testing
3714 multiple envelope elements, the last element determined the
3717 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3718 Electronic Mail") by including:
3720 Auto-submitted: auto-generated
3722 in the messages that it generates (bounce messages and others, such as
3723 warnings). In the case of bounce messages for non-SMTP messages, there was
3724 also a typo: it was using "Auto_submitted" (underscore instead of
3725 hyphen). Since every message generated by Exim is necessarily in response
3726 to another message, thes have all been changed to:
3728 Auto-Submitted: auto-replied
3730 in accordance with these statements in the RFC:
3732 The auto-replied keyword:
3734 - SHOULD be used on messages sent in direct response to another
3735 message by an automatic process,
3737 - MUST NOT be used on manually-generated messages,
3739 - MAY be used on Delivery Status Notifications (DSNs) and Message
3740 Disposition Notifications (MDNs),
3742 - MUST NOT be used on messages generated by automatic or periodic
3743 processes, except for messages which are automatic responses to
3746 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3747 to the default Received: header definition.
3749 PH/12 Added log selector acl_warn_skipped (default on).
3751 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3752 variables because (a) they are in the wrong storage pool and (b) even if
3753 they were copied, it wouldn't work properly because of the caching.
3755 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3756 checking when decoding. Apparently there are clients that generate
3757 overlong encoded strings. Why am I not surprised?
3759 PH/15 If the first argument of "${if match_address" was not empty, but did not
3760 contain an "@" character, Exim crashed. Now it writes a panic log message
3761 and treats the condition as false.
3763 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3765 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3766 extension "envelope-auth". The code is finished and in agreement with
3767 other implementations, but there is no documentation so far and in fact,
3768 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3769 not changing the active code.
3771 Print executed "if" and "elsif" statements when debugging is used. This
3772 helps a great deal to understand what a filter does.
3774 Document more things not specified clearly in RFC3028. I had all this
3775 sorted out, when out of a sudden new issues came to my mind. Oops."
3777 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3780 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3781 canonical form (as documented). However, after a host name lookup from
3782 the IP address, check_host() was doing a simple string comparison with
3783 addresses acquired from the DNS when checking that the found name did
3784 have the original IP as one of its addresses. Since any found IPv6
3785 addresses are likely to be in abbreviated form, the comparison could
3786 fail. Luckily, there already exists a function for doing the comparison
3787 by converting both addresses to binary, so now that is used instead of
3788 the text comparison.
3790 PH/20 There was another similar case to PH/19, when a complete host name was
3791 given in a host list; looking up its IP address could give an abbreviated
3792 form, whereas the current host's name might or might not be abbreviated.
3793 The same fix has been applied.
3799 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3800 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3803 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3804 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3806 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3808 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3809 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3810 Also, for "Auto-Submitted", treat anything other than "no" as
3811 non-personal, in accordance with RFC 3834. (Previously it treated
3812 anything starting "auto-" as non-personal.)
3814 TF/01 The control=submission/name=... option had a problem with syntax
3815 errors if the name included a slash character. The /name= option
3816 now slurps the rest of the string, so it can include any characters
3817 but it must come last in the list of options (after /sender_retain
3820 PH/05 Some modifications to the interface to the fake nameserver for the new
3828 TK/01 Added the "success_on_redirect" address verification option. See
3829 NewStuff for rationale and an example.
3831 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3833 PH/02 Patch to exigrep to allow it to work on syslog lines.
3835 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3836 fread() to skip over the body file's header line, because in Cygwin the
3837 header line is locked and is inaccessible.
3839 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3840 co-exist for some time) to make it clear that it is the Exim ID that is
3841 referenced, not the Message-ID: header line.
3843 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3844 string_format() function, because snprintf() does not exist on all
3847 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3848 ${stat: expansion item.
3850 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3851 protocol synchronization error", to keep the pedants happy.
3853 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3854 well as for IRIX systems, when gcc is being used. See the host.c source
3857 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3859 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3862 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3863 transport and to the smtp transport in LMTP mode.
3865 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3867 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3868 run a filter in a subprocess. This could lead to confusion in subsequent
3869 lookups in the parent process. There should also be a search_tidyup() at
3870 the end of the subprocess.
3872 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3873 only if the host matched helo_try_verify_hosts, which caused the
3874 verification to occur when the EHLO/HELO command was issued. The ACL just
3875 tested the remembered result. Now, if a previous verification attempt has
3876 not happened, "verify = helo" does it there and then.
3878 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3880 TK/03 Fix log output including CR from clamd.
3882 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3883 exist provoked a memory error which could cause a segfault.
3885 PH/15 Installed PCRE 6.2
3887 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3889 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3890 of the problem. Specifically, suggested +O2 rather than +O1 for the
3893 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3895 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3896 record, the port was not correctly logged when the outgoing_port log
3897 selector was set (it logged the transort's default port).
3899 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3900 fallback_hosts, and "hosts" in the smtp transport.
3902 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3903 host errors such as "Connection refused".
3905 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3906 authentication with radiusclient 0.4.9:
3908 - Error returned from rc_read_config was caught wrongly
3909 - Username/password not passed on to radius server due to wrong length.
3911 The presumption is that some radiusclient API changes for 4.51/PH/17
3912 were not taken care of correctly. The code is still untested by me (my
3913 Linux distribution still has 0.3.2 of radiusclient), but it was
3914 contributed by a Radius user.
3916 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3917 expanding the "port" option of the smtp transport.
3919 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3920 while reading a MIME header. Thanks to Tom Hughes for a patch.
3922 PH/24 Include config.h inside local_scan.h so that configuration settings are
3925 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3926 that in an ACL for RCPT (for example), you can examine exactly what was
3929 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3930 commands, but it was not correctly comparing the address with the actual
3931 client host address. Thus, it would show the EHLO address in Received:
3932 header lines when this was not necessary.
3934 PH/27 Added the % operator to ${eval:}.
3936 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3937 it should be ignoring failures (because with -C, for example, it has lost
3938 privilege). It wasn't ignoring creation failures other than "already
3941 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3944 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3945 because an input line was too long, either on its own, or by virtue of
3946 too many continuations, the temporary file was not being removed, and the
3947 return code was incorrect.
3949 PH/31 Missing "BOOL" in function definition in filtertest.c.
3951 PH/32 Applied Sieve patches from the maintainer.
3953 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3955 PH/33 Added "verify = not_blind".
3957 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3958 Local/Makefile (with some defaults set). These are used in built scripts
3959 such as exicyclog, but they have never been used in the exim_install
3960 script (though there are many overriding facilities there). I have
3961 arranged that the exim_install script now takes note of these two
3964 PH/35 Installed configuration files for Dragonfly.
3966 PH/36 When a locally submitted message by a trusted user did not contain a
3967 From: header, and the sender address was obtained from -f or from an SMTP
3968 MAIL command, and the trusted user did not use -F to supply a sender
3969 name, $originator_name was incorrectly used when constructing a From:
3970 header. Furthermore, $originator_name was used for submission mode
3971 messages from external hosts without From: headers in a similar way,
3972 which is clearly wrong.
3974 PH/37 Added control=suppress_local_fixups.
3976 PH/38 When log_selector = +received_sender was set, and the addition of the
3977 sender made the log line's construction buffer exactly full, or one byte
3978 less than full, an overflow happened when the terminating "\n" was
3981 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3982 when the result of a list match is failure because a DNS lookup failed.
3984 PH/40 RM_COMMAND is now used in the building process.
3986 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3987 the "build-* directories that it finds.
3989 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3990 address in a domain literal was a prefix of an interface address.
3992 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3993 when verifying a sender address, unless rewrite_headers is false.
3995 PH/44 Wrote a long comment about why errors_to addresses are verified as
3996 recipients, not senders.
3998 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3999 the ratelimit ACL was added.
4001 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4003 PH/46 Added extra information about PostgreSQL errors to the error string.
4005 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4006 avoiding the need to install special zones in a real server. This is
4007 backwards compatible; if it can't find the fake resolver, it drops back.
4008 Thus, both old and new test suites can be run.
4010 TF/02 Added util/ratelimit.pl
4012 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4013 clock is set back in time.
4015 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4016 Candler <B.Candler@pobox.com>.
4018 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4019 OK for addresses that are the result of redirections.
4021 PH/48 A number of further additions for the benefit of the new test suite,
4022 including a fake gethostbyname() that interfaces to the fake DNS resolver
4025 TF/06 The fix for widen_domains has also been applied to qualify_single and
4026 search_parents which are the other dnslookup options that can cause
4029 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4032 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4034 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4035 service" if it encountered an MX record with an empty target hostname.
4036 The message is now "an MX or SRV record indicated no SMTP service".
4038 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4039 if the DNS of the sending site is misconfigured. This is quite a
4040 common situation. This change restores the behaviour of treating a
4041 helo verification defer as a failure.
4043 PH/51 If self=fail was set on a router, the bounce message did not include the
4044 actual error message.
4050 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4052 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4053 command itself ended in error, the underlying message about the transport
4054 filter timeout was being overwritten with the pipe command error. Now the
4055 underlying error message should be appended to the second error message.
4057 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4059 PH/02 Reduce the amount of output that "make" produces by default. Full output
4060 can still be requested.
4062 PH/03 The warning log line about a condition test deferring for a "warn" verb
4063 was being output only once per connection, rather than after each
4064 occurrence (because it was using the same function as for successful
4065 "warn" verbs). This seems wrong, so I have changed it.
4067 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4068 it should not have, which might have caused a crash in the right
4069 circumstances, but probably never did.
4071 PH/04 Installed a modified version of Tony Finch's patch to make submission
4072 mode fix the return path as well as the Sender: header line, and to
4073 add a /name= option so that you can make the user's friendly name appear
4076 TF/03 Added the control = fakedefer ACL modifier.
4078 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4079 Mark Lowes for thorough testing.
4081 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4083 TK/03 Merged latest SRS patch from Miles Wilton.
4085 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4086 with the definition in sysexits.h (which is #included earlier).
4087 Fortunately, Exim does not actually use EX_OK. The code used to try to
4088 preserve the sysexits.h value, by assuming that macro definitions were
4089 scanned for macro replacements. I have been disabused of this notion,
4090 so now the code just undefines EX_OK before #including unistd.h.
4092 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4093 in the smtp transport. When a block could not be written in a single
4094 write() function, the timeout was being re-applied to each part-write.
4095 This seems wrong - if the receiver was accepting one byte at a time it
4096 would take for ever. The timeout is now adjusted when this happens. It
4097 doesn't have to be particularly precise.
4099 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4100 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4102 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4103 without a domain if the check to <postmaster@domain> fails.
4105 SC/01 Eximstats: added -xls and the ability to specify output files
4106 (patch written by Frank Heydlauf).
4108 SC/02 Eximstats: use FileHandles for outputting results.
4110 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4112 SC/04 Eximstats: fixed display of large numbers with -nvr option
4114 SC/05 Eximstats: fixed merging of reports with empty tables.
4116 SC/06 Eximstats: added the -include_original_destination flag
4118 SC/07 Eximstats: removed tabs and trailing whitespace.
4120 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4122 TK/06 MBOX spool code: Add real "From " MBOX separator line
4123 so the .eml file is really in mbox format (even though
4124 most programs do not really care). Patch from Alex Miller.
4126 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4127 The latter is generated from $received_to and is only set if the
4128 message has one envelope recipient. SA can use these headers,
4129 obviously out-of-the-box. Patch from Alex Miller.
4131 PH/08 The ${def test on a variable was returning false if the variable's
4132 value was "0", contrary to what the specification has always said!
4133 The result should be true unless the variable is empty.
4135 PH/09 The syntax error of a character other than { following "${if
4136 def:variable_name" (after optional whitespace) was not being diagnosed.
4137 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4138 accidental colon was present, for example, could give incorrect results.
4140 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4141 result is used (not including appendfile, where other changes are about
4144 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4145 This involved changing a lot of size variables from int to off_t. It
4146 should work with maildirs and everything.
4148 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4149 spamd dying while we are connected to it.
4151 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4154 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4155 function for BDB 4.3.
4157 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4159 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4160 the sources for PCRE have been re-arranged and more files are now
4163 PH/15 The code I had for printing potentially long long variables in PH/11
4164 above was not the best (it lost precision). The length of off_t variables
4165 is now inspected at build time, and an appropriate printing format (%ld
4166 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4167 to be "long long int" or "long int". This is needed for the internal
4168 formatting function string_vformat().
4170 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4171 the configuration file to be ":syslog", then the script "guesses" where
4172 the logs files are, rather than using the compiled in default. In our
4173 case the guess is not the same as the compiled default, so the script
4174 suddenly stopped working when I started to use syslog. The patch checks
4175 to see if log_file_path is "". If so, it attempts to read it from exim
4176 with no configuration file to get the compiled in version, before it
4177 falls back to the previous guessing code."
4179 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4180 implementing BATV in an Exim configuration. See NewStuff for the gory
4183 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4184 Makefile that are specific to HP-UX.
4186 PH/18 If the "use_postmaster" option was set for a recipient callout together
4187 with the "random" option, the postmaster address was used as the MAIL
4188 FROM address for the random test, but not for the subsequent recipient
4189 test. It is now used for both.
4191 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4192 patch removes a few documentation additions to RFC 3028, because the
4193 latest draft now contains them. It adds the new en;ascii-case comparator
4194 and a new error check for 8bit text in MIME parts. Comparator and
4195 require names are now matched exactly. I enabled the subaddress
4196 extension, but it is not well tested yet (read: it works for me)."
4198 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4199 rework some of the code of TK/09 above to avoid the hardwired use of
4200 "%lld" and "long long". Replaced the call to snprintf() with a call to
4203 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4204 records point to non-existent hosts", "retry timeout exceeded", and
4205 "retry time not reached for any host after a long failure period".
4207 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4208 experimental DomainKeys support:
4210 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4211 (2) On an error such as an illegally used "control", the wrong name for
4212 the control was given.
4214 These problems did NOT occur unless DomainKeys support was compiled.
4216 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4218 PH/24 Added ${if match_ip condition.
4220 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4221 databases so that it will be absolutely obvious if a crash occurs in the
4222 DB library. This is a regular occurrence (often caused by mis-matched
4225 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4226 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4227 picked up on a user's system that detects such things. There doesn't seem
4228 to be a gcc warning option for this - only an attribute that has to be
4229 put on the function's prototype. It seems that in Fedora Core 4 they have
4230 set this on a number of new functions. No doubt there will be more in due
4233 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4234 specify a transport. However, if an address that was verified by such a
4235 router was the subject of a callout, Exim crashed because it tried to
4236 read the rcpt_include_affixes from the non-existent transport. Now it
4237 just assumes that the setting of that option is false. This bug was
4238 introduced by 4.51/PH/31.
4240 PH/28 Changed -d+all to exclude +memory, because that information is very
4241 rarely of interest, but it makes the output a lot bigger. People tend to
4242 do -d+all out of habit.
4244 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4245 code in os-type was giving problems when libc.so lives in lib64, like on
4248 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4249 aren't the modern standard, and it seems that some systems' include files
4250 don't always have them. Exim was already checking for some of the newer
4251 ones like T_AAAA, and defining it itself. I've added checks for all the
4252 record types that Exim uses.
4254 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4255 not automatically generating a new one, as it is supposed to. This
4256 prevented TLS from working. If the file did exist, but contained invalid
4257 data, a new version was generated, as expected. It was only the case of a
4258 non-existent file that was broken.
4260 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4261 with a change in libdomainkeys > 0.64.
4263 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4264 from DNS. If the selector record carries the flag, it now has
4265 precedence over the domain-wide flag.
4267 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4269 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4270 the use of a port name that isn't defined in /etc/services) occurred, the
4271 message was deferred as in a normal delivery, and thus remained on the
4272 spool, instead of being failed because of the mua_wrapper setting. This
4273 is now fixed, and I tidied up some of the mua_wrapper messages at the
4276 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4277 the messages in a hash of arrays rather than using individual hashes.
4278 This is a bit cleaner and results in dramatic memory savings, albeit
4279 at a slight CPU cost.
4281 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4282 as requested by Marc Sherman.
4284 SC/10 Eximstats: added histograms for user specified patterns as requested
4287 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4289 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4290 fopen() in the content-scanning modules that did not already have it.
4296 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4297 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4299 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4301 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4303 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4304 to test Sieve filters that use "vacation".
4306 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4307 that changes the way the GnuTLS parameters are stored in the cache file.
4308 The new format can be generated externally. For backward compatibility,
4309 if the data in the cache doesn't make sense, Exim assumes it has read an
4310 old-format file, and it generates new data and writes a new file. This
4311 means that you can't go back to an older release without removing the
4314 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4315 work if there are any delivery delays because "one_time" forces the
4316 parent to be marked "delivered", so its unseen clone is never tried
4317 again. For this reason, Exim now forbids the simultaneous setting of
4320 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4321 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4322 entry if you want to know the details. The fix, however, neglected to
4323 consider the case when local delivery batching is involved. The test for
4324 "previously delivered" was not happening when checking to see if an
4325 address could be batched with a previous (undelivered) one; under
4326 certain circumstances this could lead to multiple deliveries to the same
4329 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4330 in its include files, and this causes problems building Exim.
4332 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4333 header_syntax) but Exim was just ignoring anything given after a slash.
4334 In particular, this caused confusion with an attempt to use "verify =
4335 reverse_host_lookup/defer_ok". An error is now given when options are
4336 supplied for verify items that do not have them. (Maybe reverse_host_
4337 lookup should have a defer_ok option, but that's a different point.)
4339 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4340 defined by RFC 821) to 2048, because there were problems with some AUTH
4341 commands, and RFC 1869 says the size should be increased for extended
4342 SMTP commands that take arguments.
4344 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4347 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4348 "unknown" error; now it says that the functionality isn't in the binary.
4350 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4351 an address' error message when a string expansion fails (syntax or
4352 whatever). Otherwise the password may appear in the log. Following change
4353 PH/42 below, there is no longer a chance of it appearing in a bounce
4356 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4358 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4359 of its addresses was ever tried. (Bugzilla bug #2.)
4361 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4362 the result incorrectly in the debug output. (It correctly added a newline
4363 to what was transported.)
4365 TF/01 Added $received_time.
4367 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4368 commented out examples of how to interface to a virus scanner and to
4369 SpamAssassin. Also added commented examples of av_scanner and
4370 spamd_address settings.
4372 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4373 and controls are allowed in which ACLs. There were a couple of minor
4374 errors. Some of the entries in the conditions table (which is a table of
4375 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4376 negation of where the condition IS allowed.
4378 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4380 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4381 header file does not have a version number, so I've had to invent a new
4382 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4383 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4384 radiusclient), but it was contributed by a Radius user.
4386 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4387 files or directories, for OpenSSL.
4389 PH/19 When an Exim process that is running as root has to create an Exim log
4390 file, it does so in a subprocess that runs as exim:exim so as to get the
4391 ownership right at creation (otherwise, other Exim processes might see
4392 the file with the wrong ownership). There was no test for failure of this
4393 fork() call, which would lead to the process getting stuck as it waited
4394 for a non-existent subprocess. Forks do occasionally fail when resources
4395 run out. I reviewed all the other calls to fork(); they all seem to check
4398 PH/20 When checking for unexpected SMTP input at connect time (before writing
4399 the banner), Exim was not dealing correctly with a non-positive return
4400 from the read() function. If the client had disconnected by this time,
4401 the result was a log entry for a synchronization error with an empty
4402 string after "input=" when read() returned zero. If read() returned -1
4403 (an event I could not check), uninitialized data bytes were printed.
4404 There were reports of junk text (parts of files, etc) appearing after
4407 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4409 PH/22 Added support for macro redefinition, and (re)definition in between
4410 driver and ACL definitions.
4412 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4413 forgetting to use the resulting value; it was using the unexpanded value.
4415 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4416 hadn't been configured. The fix is from Juergen Kreileder, who
4417 understands it better than I do:
4419 "Here's what I see happening with three configured cyrus_sasl
4420 authenticators configured (plain, login, cram-md5):
4422 On startup auth_cyrus_sasl_init() gets called for each of these.
4423 This means three calls to sasl_listmech() without a specified mech_list.
4424 => SASL tests which mechs of all available mechs actually work
4425 => three warnings about OTP not working
4426 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4428 With the patch, sasl_listmech() also gets called three times. But now
4429 SASL's mech_list option is set to the server_mech specified in the the
4430 authenticator. Or in other words, the answer from sasl_listmech()
4431 gets limited to just the mech you're testing for (which is different
4433 => the return list contains just 'plain' or 'login', 'cram-md5' or
4434 nothing depending on the value of ob->server_mech.
4436 I've just tested the patch: Authentication still works fine,
4437 unavailable mechs specified in the exim configuration are still
4438 caught, and the auth.log warnings about OTP are gone."
4440 PH/25 When debugging is enabled, the contents of the command line are added
4441 to the debugging output, even when log_selector=+arguments is not
4444 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4445 answer is "GNU", and only if the return is "GNU/something" is the answer
4448 PH/27 $acl_verify_message is now set immediately after the failure of a
4449 verification in an ACL, and so is available in subsequent modifiers. In
4450 particular, the message can be preserved by coding like this:
4452 warn !verify = sender
4453 set acl_m0 = $acl_verify_message
4455 Previously, $acl_verify_message was set only while expanding "message"
4456 and "log_message" when a very denied access.
4458 PH/28 Modified OS/os.c-Linux with
4460 -#ifndef OS_LOAD_AVERAGE
4461 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4463 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4464 nomenclature these days.)
4466 PH/29 Installed patch from the Sieve maintainer that adds the options
4467 sieve_useraddress and sieve_subaddress to the redirect router.
4469 PH/30 In these circumstances:
4470 . Two addresses routed to the same list of hosts;
4471 . First host does not offer TLS;
4472 . First host accepts first address;
4473 . First host gives temporary error to second address;
4474 . Second host offers TLS and a TLS session is established;
4475 . Second host accepts second address.
4476 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4477 and peerdn, if requested) that were in fact used only for the second
4480 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4481 attention to any local part prefix or suffix that was matched by the
4482 router that accepted the address. It now behaves in the same way as it
4483 does for delivery: the affixes are removed from the local part unless
4484 rcpt_include_affixes is set on the transport.
4486 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4487 timeout during the DATA phase of an incoming message.
4489 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4490 applied a patch sanctioned by the Sieve maintainer.
4492 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4493 the uid or gid is negative. A case of a negative gid caused this to be
4494 noticed. The fix allows for either to be negative.
4496 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4497 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4500 PH/36 The change PH/12 above was broken. Fixed it.
4502 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4503 the grounds that routing the same address twice would always produce the
4504 same answer. This might have been true once, but it is certainly no
4505 longer true now. Routing a child address may depend on the previous
4506 routing that produced that child. Some complicated redirection strategies
4507 went wrong when messages had multiple recipients, and made Exim's
4508 behaviour dependent on the order in which the addresses were given.
4510 I have moved the duplicate checking until after the routing is complete.
4511 Exim scans the addresses that are assigned to local and remote
4512 transports, and removes any duplicates. This means that more work will be
4513 done, as duplicates will always all be routed, but duplicates are
4514 presumably rare, so I don't expect this is of any significance.
4516 For deliveries to pipes, files, and autoreplies, the duplicate checking
4517 still happens during the routing process, since they are not going to be
4520 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4521 It corrects a timeout issue with spamd. This is Ian's comment: "The
4522 background is that sometimes spamd either never reads data from a
4523 connection it has accepted, or it never writes response data. The exiscan
4524 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4525 blindly assumes that writes won't block so it may never time out."
4527 PH/39 Allow G after quota size as well as K and M.
4529 PH/40 The value set for $authenticated_id in an authenticator may not contain
4530 binary zeroes or newlines because the value is written to log lines and
4531 to spool files. There was no check on this. Now the value is run through
4532 the string_printing() function so that such characters are converted to
4533 printable escape sequences.
4535 PH/41 $message_linecount is a new variable that contains the total number of
4536 lines in the message. Compare $body_linecount, which is the count for the
4539 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4540 bounce and delay warning messages, except in certain special cases, which
4543 (a) An SMTP error message from a remote host;
4544 (b) A message specified in a :fail: redirection;
4545 (c) A message specified in a "fail" command in a system filter;
4546 (d) A message specified in a FAIL return from the queryprogram router;
4547 (e) A message specified by the cannot_route_message router option.
4549 In these cases only, Exim does include the error details in bounce and
4550 warning messages. There are also a few cases where bland messages such
4551 as "unrouteable address" or "local delivery error" are given.
4553 PH/43 $value is now also set for the "else" part of a ${run expansion.
4555 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4556 being worked on, but at least Exim now implements the latest version to
4559 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4560 process to complete was treated as a delivery failure, a timeout while
4561 writing the message to the pipe was logged, but erroneously treated as a
4562 successful delivery. Such timeouts include transport filter timeouts. For
4563 consistency with the overall process timeout, these timeouts are now
4564 treated as errors, giving rise to delivery failures by default. However,
4565 there is now a new Boolean option for the pipe transport called
4566 timeout_defer, which, if set TRUE, converts the failures into defers for
4567 both kinds of timeout. A transport filter timeout is now identified in
4570 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4571 systems where "make" and "gmake" are different, calling "gmake" at top
4572 level broke things. I've arranged for the value of $(MAKE) to be passed
4573 from the Makefile to this script so that it can call the same version of
4577 A note about Exim versions 4.44 and 4.50
4578 ----------------------------------------
4580 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4581 changes of various kinds. As a consequence, a big documentation update was
4582 needed. This delayed the release for rather longer than seemed good, especially
4583 in the light of a couple of (minor) security issues. Therefore, the changes
4584 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4585 release. So 4.44 and 4.50 are in effect two different branches that both start
4588 I have left the 4.50 change log unchanged; it contains all the changes since
4589 4.43. The change log for 4.44 is below; many of its items are identical to
4590 those for 4.50. This seems to be the most sensible way to preserve the
4591 historical information.
4597 1. Minor wording change to the doc/README.SIEVE file.
4599 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4600 computation of the current number of files was incorrect.
4602 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4603 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4606 4. Give more explanation in the error message when the command for a transport
4607 filter fails to execute.
4609 5. There are several places where Exim runs a non-Exim command in a
4610 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4611 was being done only for the command run by the queryprogram router. It is
4612 now done for all such subprocesses. The other cases are: ${run, transport
4613 filters, and the commands run by the lmtp and pipe transports.
4615 6. Added CONFIGURE_GROUP build-time option.
4617 7. Some older OS have a limit of 256 on the maximum number of file
4618 descriptors. Exim was using setrlimit() to set 1000 as a large value
4619 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4620 systems. I've change it so that if it can't get 1000, it tries for 256.
4622 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4623 was an oversight, and furthermore, ever since the addition of extra
4624 controls (e.g. 4.43/32), the checks on when to allow different forms of
4625 "control" were broken. There should now be diagnostics for all cases when a
4626 control that does not make sense is encountered.
4628 9. Added the /retain_sender option to "control=submission".
4630 10. $recipients is now available in the predata ACL (oversight).
4632 11. Tidy the search cache before the fork to do a delivery from a message
4633 received from the command line. Otherwise the child will trigger a lookup
4634 failure and thereby defer the delivery if it tries to use (for example) a
4635 cached ldap connection that the parent has called unbind on.
4637 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4638 of $address_data from the recipient verification was clobbered by the
4639 sender verification.
4641 13. The value of address_data from a sender verification is now available in
4642 $sender_address_data in subsequent conditions in the ACL statement.
4644 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4646 15. Added a new option "connect=<time>" to callout options, to set a different
4649 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4650 was its contents. (It was OK if the option was not defined at all.)
4652 17. A "Completed" log line is now written for messages that are removed from
4653 the spool by the -Mrm option.
4655 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4656 information about exactly what failed.
4658 19. Added -dd to debug only the daemon process.
4660 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4661 handles timeouts, both on the server side and network timeouts. Renamed the
4662 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4664 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4665 It is now set to "smtps".
4667 22. $host_address is now set to the target address during the checking of
4668 ignore_target_hosts.
4670 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4671 being passed; this would have caused $sender_host_name to have been used if
4672 matching the list had actually called for a host name (not very likely,
4673 since this list is usually IP addresses). A host name is now passed as
4676 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4677 code that specifies a non-restarting handler (typically sigaction() in
4678 modern systems) in an attempt to fix a rare and obscure crash bug.
4680 25. Narrowed the window for a race in the daemon that could cause it to ignore
4681 SIGCHLD signals. This is not a major problem, because they are used only to
4682 wake it up if nothing else does.
4684 26. A malformed maildirsize file could cause Exim to calculate negative values
4685 for the mailbox size or file count. Odd effects could occur as a result.
4686 The maildirsize information is now recalculated if the size or filecount
4689 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4690 support for a long time. Removed HAVE_SYS_VFS_H.
4692 28. Installed the latest version of exipick from John Jetmore.
4694 29. In an address list, if the pattern was not a regular expression, an empty
4695 subject address (from a bounce message) matched only if the pattern was an
4696 empty string. Non-empty patterns were not even tested. This was the wrong
4697 because it is perfectly reasonable to use an empty address as part of a
4698 database query. An empty address is now tested by patterns that are
4699 lookups. However, all the other forms of pattern expect the subject to
4700 contain a local part and a domain, and therefore, for them, an empty
4701 address still always fails if the pattern is not itself empty.
4703 30. Exim went into a mad DNS loop when attempting to do a callout where the
4704 host was specified on an smtp transport, and looking it up yielded more
4705 than one IP address.
4707 31. Re-factored the code for checking spool and log partition space into a
4708 function that finds that data and another that does the check. The former
4709 is then used to implement four new variables: $spool_space, $log_space,
4710 $spool_inodes, and $log_inodes.
4712 32. The RFC2047 encoding function was originally intended for short strings
4713 such as real names; it was not keeping to the 75-character limit for
4714 encoded words that the RFC imposes. It now respects the limit, and
4715 generates multiple encoded words if necessary. To be on the safe side, I
4716 have increased the buffer size for the ${rfc2047: expansion operator from
4719 33. It is now permitted to omit both strings after an "if" condition; if the
4720 condition is true, the result is "true". As before, when the second string
4721 is omitted, a false condition yields an empty string. This makes it less
4722 cumbersome to write custom ACL and router conditions.
4724 34. Failure to deliver a bounce message always caused it to be frozen, even if
4725 there was an errors_to setting on the router. The errors_to setting is now
4728 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4729 canonical form (fully expanded) before being placed in
4730 $sender_host_address.
4732 36. The table in the code that translates DNS record types into text (T_A to
4733 "A" for instance) was missing entries for NS and CNAME. It is just possible
4734 that this could have caused confusion if both these types were looked up
4735 for the same domain, because the text type is used as part of Exim's
4736 per-process caching. But the chance of anyone hitting this buglet seems
4739 37. The dnsdb lookup has been extended in a number of ways.
4741 (1) There is a new type, "zns", which walks up the domain tree until it
4742 finds some nameserver records. It should be used with care.
4744 (2) There is a new type, "mxh", which is like "mx" except that it returns
4745 just the host names, not the priorities.
4747 (3) It is now possible to give a list of domains (or IP addresses) to be
4748 looked up. The behaviour when one of the lookups defers can be
4749 controlled by a keyword.
4751 (4) It is now possible to specify the separator character for use when
4752 multiple records are returned.
4754 38. The dnslists ACL condition has been extended: it is now possible to supply
4755 a list of IP addresses and/or domains to be looked up in a particular DNS
4758 39. Added log_selector=+queue_time_overall.
4760 40. When running the queue in the test harness, wait just a tad after forking a
4761 delivery process, to get repeatability of debugging output.
4763 41. Include certificate and key file names in error message when GnuTLS fails
4764 to set them up, because the GnuTLS error message doesn't include the name
4765 of the failing file when there is a problem reading it.
4767 42. Allow both -bf and -bF in the same test run.
4769 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4771 44. The "Exiscan patch" is now merged into the mainline Exim source.
4773 45. Sometimes the final signoff response after QUIT could fail to get
4774 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4775 < 0 before doing a fflush(). This bug looks as though it goes back to the
4776 introduction of TLS in release 3.20, but "sometimes" must have been rare
4777 because the tests only now provoked it.
4779 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4780 (this can affect the format of dates).
4782 47. exim_tidydb, when checking for the continued existence of a message for
4783 which it has found a message-specific retry record, was not finding
4784 messages that were in split spool directories. Consequently, it was
4785 deleting retry records that should have stayed in existence.
4787 48. Steve fixed some bugs in eximstats.
4789 49. The SPA authentication driver was not abandoning authentication and moving
4790 on to the next authenticator when an expansion was forced to fail,
4791 contradicting the general specification for all authenticators. Instead it
4792 was generating a temporary error. It now behaves as specified.
4794 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4795 (the order specifies the preference for clients). The order is now AES256,
4796 AES128, 3DES, ARCFOUR128.
4798 51. Small patch to Sieve code - explicitly set From: when generating an
4801 52. Exim crashed if a remote delivery caused a very long error message to be
4802 recorded - for instance if somebody sent an entire SpamAssassin report back
4803 as a large number of 550 error lines. This bug was coincidentally fixed by
4804 increasing the size of one of Exim's internal buffers (big_buffer) that
4805 happened as part of the Exiscan merge. However, to be on the safe side, I
4806 have made the code more robust (and fixed the comments that describe what
4809 53. Now that there can be additional text after "Completed" in log lines (if
4810 the queue_time_overall log selector is set), a one-byte patch to exigrep
4811 was needed to allow it to recognize "Completed" as not the last thing in
4814 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4815 patch that reportedly fixes this has been added. I am not expert enough to
4816 create a test for it. This is what the patch creator wrote:
4818 "I found a little strange behaviour of ldap code when working with
4819 Windows 2003 AD Domain, where users was placed in more than one
4820 Organization Units. When I tried to give exim partial DN, the exit code
4821 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4822 But simultaneously result of request was absolutely normal ldap result,
4823 so I produce this patch..."
4825 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4826 REFERENCE, so I have modified the code to exclude the patch when that macro
4829 55. Some experimental protocols are using DNS PTR records for new purposes. The
4830 keys for these records are domain names, not reversed IP addresses. The
4831 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4832 leaves it alone. Component reversal etc. now happens only for IP addresses.
4835 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4837 57. Double the size of the debug message buffer (to 2048) so that more of very
4838 long debug lines gets shown.
4840 58. The exicyclog utility now does better if the number of log files to keep
4841 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4843 59. Two changes related to the smtp_active_hostname option:
4845 (1) $smtp_active_hostname is now available as a variable.
4846 (2) The default for smtp_banner uses $smtp_active_hostname instead
4847 of $primary_hostname.
4849 60. The host_aton() function is supposed to be passed a string that is known
4850 to be a valid IP address. However, in the case of IPv6 addresses, it was
4851 not checking this. This is a hostage to fortune. Exim now panics and dies
4852 if the condition is not met. A case was found where this could be provoked
4853 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4854 components; fortuitously, this particular loophole had already been fixed
4855 by change 4.50/55 above.
4857 If there are any other similar loopholes, the new check in host_aton()
4858 itself should stop them being exploited. The report I received stated that
4859 data on the command line could provoke the exploit when Exim was running as
4860 exim, but did not say which command line option was involved. All I could
4861 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4862 running as the user.
4865 61. There was a buffer overflow vulnerability in the SPA authentication code
4866 (which came originally from the Samba project). I have added a test to the
4867 spa_base64_to_bits() function which I hope fixes it.
4870 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4871 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4873 63. The daemon start-up calls getloadavg() while still root for those OS that
4874 need the first call to be done as root, but it missed one case: when
4875 deliver_queue_load_max is set with deliver_drop_privilege. This is
4876 necessary for the benefit of the queue runner, because there is no re-exec
4877 when deliver_drop_privilege is set.
4879 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4880 This has been fixed.
4882 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4883 was in use, was not putting the data itself into the right store pool;
4884 consequently, it could be overwritten for a subsequent message in the same
4885 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4888 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4890 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4891 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4892 the function were treating the return as a boolean value, which happened to
4893 work because 0=false and not-0=true, but is not correct code.
4895 68. The host_aton() function was not handling scoped IPv6 addresses (those
4896 with, for example, "%eth0" on the end) correctly.
4898 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4899 negated items (that is, ~something) in unsigned ints. Some compilers
4900 apparently mutter when there is no cast.
4902 70. If an address verification called from an ACL failed, and did not produce a
4903 user-specific message (i.e. there was only a "system" message), nothing was
4904 put in $acl_verify_message. In this situation, it now puts the system
4907 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4908 session; change 4.31/43 added the unwanted input to the log line - except
4909 that it did not do this in the start of session case. It now does.
4911 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4912 This is wrong and can cause the other end to generate a synchronization
4913 error if it is another Exim or anything else that does the synchronization
4914 check. A QUIT command is no longer sent after a timeout.
4916 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4917 during host lookups.
4919 74. The defer_ok option of callout verification was not working if it was used
4920 when verifying addresses in header lines, that is, for this case:
4922 verify = header_sender/callout=defer_ok
4924 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4925 those file descriptors could be used for SMTP connections. If anything
4926 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4927 could be sent to the SMTP client, causing chaos. The daemon now opens
4928 stdin, stdout, and stderr to /dev/null when it puts itself into the
4931 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4932 log by default. The user can override this with suitable Perl magic.
4934 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4935 the log message when discard triggers, was not working for the DATA ACL or
4936 for the non-SMTP ACL.
4938 78. Error message wording change in sieve.c.
4940 79. If smtp_accept_max_per_host was set, the number of connections could be
4941 restricted to fewer than expected, because the daemon was trying to set up
4942 a new connection before checking whether the processes handling previous
4943 connections had finished. The check for completed processes is now done
4944 earlier. On busy systems, this bug wouldn't be noticed because something
4945 else would have woken the daemon, and it would have reaped the completed
4948 80. If a message was submitted locally by a user whose login name contained one
4949 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4950 It caused a spool format error. I have fixed the spool reading code. A
4951 related problem was that the "from" clause in the Received: line became
4952 illegal because of the space(s). It is now covered by ${quote_local_part.
4954 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4957 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4959 83. Patch from Sieve maintainer to fix unterminated string problem in
4960 "vacation" handling.
4962 84. Some minor changes to the Linux configuration files to help with other
4963 OS variants using glibc.
4965 85. One more patch for Sieve to update vacation handling to latest spec.
4968 ----------------------------------------------------
4969 See the note above about the 4.44 and 4.50 releases.
4970 ----------------------------------------------------
4976 1. Change 4.43/35 introduced a bug that caused file counts to be
4977 incorrectly computed when quota_filecount was set in an appendfile
4980 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4981 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4984 3. Give more explanation in the error message when the command for a transport
4985 filter fails to execute.
4987 4. There are several places where Exim runs a non-Exim command in a
4988 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4989 was being done only for the command run by the queryprogram router. It is
4990 now done for all such subprocesses. The other cases are: ${run, transport
4991 filters, and the commands run by the lmtp and pipe transports.
4993 5. Some older OS have a limit of 256 on the maximum number of file
4994 descriptors. Exim was using setrlimit() to set 1000 as a large value
4995 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4996 systems. I've change it so that if it can't get 1000, it tries for 256.
4998 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4999 was an oversight, and furthermore, ever since the addition of extra
5000 controls (e.g. 4.43/32), the checks on when to allow different forms of
5001 "control" were broken. There should now be diagnostics for all cases when a
5002 control that does not make sense is encountered.
5004 7. $recipients is now available in the predata ACL (oversight).
5006 8. Tidy the search cache before the fork to do a delivery from a message
5007 received from the command line. Otherwise the child will trigger a lookup
5008 failure and thereby defer the delivery if it tries to use (for example) a
5009 cached ldap connection that the parent has called unbind on.
5011 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5012 of $address_data from the recipient verification was clobbered by the
5013 sender verification.
5015 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5016 was its contents. (It was OK if the option was not defined at all.)
5018 11. A "Completed" log line is now written for messages that are removed from
5019 the spool by the -Mrm option.
5021 12. $host_address is now set to the target address during the checking of
5022 ignore_target_hosts.
5024 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5025 being passed; this would have caused $sender_host_name to have been used if
5026 matching the list had actually called for a host name (not very likely,
5027 since this list is usually IP addresses). A host name is now passed as
5030 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5031 code that specifies a non-restarting handler (typically sigaction() in
5032 modern systems) in an attempt to fix a rare and obscure crash bug.
5034 15. Narrowed the window for a race in the daemon that could cause it to ignore
5035 SIGCHLD signals. This is not a major problem, because they are used only to
5036 wake it up if nothing else does.
5038 16. A malformed maildirsize file could cause Exim to calculate negative values
5039 for the mailbox size or file count. Odd effects could occur as a result.
5040 The maildirsize information is now recalculated if the size or filecount
5043 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5044 support for a long time. Removed HAVE_SYS_VFS_H.
5046 18. Updated exipick to current release from John Jetmore.
5048 19. Allow an empty sender to be matched against a lookup in an address list.
5049 Previously the only cases considered were a regular expression, or an
5052 20. Exim went into a mad DNS lookup loop when doing a callout where the
5053 host was specified on the transport, if the DNS lookup yielded more than
5056 21. The RFC2047 encoding function was originally intended for short strings
5057 such as real names; it was not keeping to the 75-character limit for
5058 encoded words that the RFC imposes. It now respects the limit, and
5059 generates multiple encoded words if necessary. To be on the safe side, I
5060 have increased the buffer size for the ${rfc2047: expansion operator from
5063 22. Failure to deliver a bounce message always caused it to be frozen, even if
5064 there was an errors_to setting on the router. The errors_to setting is now
5067 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5068 canonical form (fully expanded) before being placed in
5069 $sender_host_address.
5071 24. Updated eximstats to version 1.33
5073 25. Include certificate and key file names in error message when GnuTLS fails
5074 to set them up, because the GnuTLS error message doesn't include the name
5075 of the failing file when there is a problem reading it.
5077 26. Expand error message when OpenSSL has problems setting up cert/key files.
5080 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5081 (this can affect the format of dates).
5083 28. exim_tidydb, when checking for the continued existence of a message for
5084 which it has found a message-specific retry record, was not finding
5085 messages that were in split spool directories. Consequently, it was
5086 deleting retry records that should have stayed in existence.
5088 29. eximstats updated to version 1.35
5089 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5090 1.35 - bugfix such that pie charts by volume are generated correctly
5092 30. The SPA authentication driver was not abandoning authentication and moving
5093 on to the next authenticator when an expansion was forced to fail,
5094 contradicting the general specification for all authenticators. Instead it
5095 was generating a temporary error. It now behaves as specified.
5097 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5098 (the order specifies the preference for clients). The order is now AES256,
5099 AES128, 3DES, ARCFOUR128.
5101 31. Small patch to Sieve code - explicitly set From: when generating an
5104 32. Exim crashed if a remote delivery caused a very long error message to be
5105 recorded - for instance if somebody sent an entire SpamAssassin report back
5106 as a large number of 550 error lines. This bug was coincidentally fixed by
5107 increasing the size of one of Exim's internal buffers (big_buffer) that
5108 happened as part of the Exiscan merge. However, to be on the safe side, I
5109 have made the code more robust (and fixed the comments that describe what
5112 33. Some experimental protocols are using DNS PTR records for new purposes. The
5113 keys for these records are domain names, not reversed IP addresses. The
5114 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5115 leaves it alone. Component reversal etc. now happens only for IP addresses.
5118 34. The host_aton() function is supposed to be passed a string that is known
5119 to be a valid IP address. However, in the case of IPv6 addresses, it was
5120 not checking this. This is a hostage to fortune. Exim now panics and dies
5121 if the condition is not met. A case was found where this could be provoked
5122 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5123 components; fortuitously, this particular loophole had already been fixed
5124 by change 4.50/55 or 4.44/33 above.
5126 If there are any other similar loopholes, the new check in host_aton()
5127 itself should stop them being exploited. The report I received stated that
5128 data on the command line could provoke the exploit when Exim was running as
5129 exim, but did not say which command line option was involved. All I could
5130 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5131 running as the user.
5134 35. There was a buffer overflow vulnerability in the SPA authentication code
5135 (which came originally from the Samba project). I have added a test to the
5136 spa_base64_to_bits() function which I hope fixes it.
5139 36. The daemon start-up calls getloadavg() while still root for those OS that
5140 need the first call to be done as root, but it missed one case: when
5141 deliver_queue_load_max is set with deliver_drop_privilege. This is
5142 necessary for the benefit of the queue runner, because there is no re-exec
5143 when deliver_drop_privilege is set.
5145 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5146 was in use, was not putting the data itself into the right store pool;
5147 consequently, it could be overwritten for a subsequent message in the same
5148 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5151 38. Sometimes the final signoff response after QUIT could fail to get
5152 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5153 < 0 before doing a fflush(). This bug looks as though it goes back to the
5154 introduction of TLS in release 3.20, but "sometimes" must have been rare
5155 because the tests only now provoked it.
5161 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5162 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5163 Now it returns an integer. A number of calls were still expecting a T/F
5164 return. Fortuitously, in all cases, the tests worked in OK situations,
5165 which is the norm. However, things would have gone wrong on any write
5166 failures on the smtp file descriptor. This function is used when sending
5167 messages over SMTP and also when doing verify callouts.
5169 2. When Exim is called to do synchronous delivery of a locally submitted
5170 message (the -odf or -odi options), it no longer closes stderr before doing
5173 3. Implemented the mua_wrapper option.
5175 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5177 5. Implemented the functions header_remove(), header_testname(),
5178 header_add_at_position(), and receive_remove_recipient(), and exported them
5181 6. If an ACL "warn" statement specified the addition of headers, Exim already
5182 inserted X-ACL-Warn: at the start if there was no header name. However, it
5183 was not making this test for the second and subsequent header lines if
5184 there were newlines in the string. This meant that an invalid header could
5185 be inserted if Exim was badly configured.
5187 7. Allow an ACL "warn" statement to add header lines at the start or after all
5188 the Received: headers, as well as at the end.
5190 8. Added the rcpt_4xx retry error code.
5192 9. Added postmaster_mailfrom=xxx to callout verification option.
5194 10. Added mailfrom=xxxx to the callout verification option, for verify=
5197 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5198 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5200 12. Inserted some casts to stop certain compilers warning when using pointer
5201 differences as field lengths or precisions in printf-type calls (mostly
5202 affecting debugging statements).
5204 13. Added optional readline() support for -be (dynamically loaded).
5206 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5207 same clock tick as a message's arrival, so that its received time was the
5208 same as the "first fail" time on the retry record, and that message
5209 remained on the queue past the ultimate address timeout, every queue runner
5210 would try a delivery (because it was past the ultimate address timeout) but
5211 after another failure, the ultimate address timeout, which should have then
5212 bounced the address, did not kick in. This was a "< instead of <=" error;
5213 in most cases the first failure would have been in the next clock tick
5214 after the received time, and all would be well.
5216 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5217 being recognized when the domain list was tested by the match_domain
5218 condition in an expansion string.
5220 16. Added the ${str2b64: operator.
5222 17. Exim was always calling setrlimit() to set a large limit for the number of
5223 processes, without checking whether the existing limit was already
5224 adequate. (It did check for the limit on file descriptors.) Furthermore,
5225 errors from getrlimit() and setrlimit() were being ignored. Now they are
5226 logged to the main and panic logs, but Exim does carry on, to try to do its
5227 job under whatever limits there are.
5229 18. Imported PCRE 5.0.
5231 19. Trivial typo in log message " temporarily refused connection" (the leading
5234 20. If the log selector return_path_on_delivery was set and an address was
5235 redirected to /dev/null, the delivery process crashed because it assumed
5236 that a return path would always be set for a "successful" delivery. In this
5237 case, the whole delivery is bypassed as an optimization, and therefore no
5240 21. Internal re-arrangement: the function for sending a challenge and reading
5241 a response while authentication was assuming a zero-terminated challenge
5242 string. It's now changed to take a pointer and a length, to allow for
5243 binary data in such strings.
5245 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5247 23. Exim was not respecting finduser_retries when seeking the login of the
5248 uid under which it was called; it was always trying 10 times. (The default
5249 setting of finduser_retries is zero.) Also, it was sleeping after the final
5250 failure, which is pointless.
5252 24. Implemented tls_on_connect_ports.
5254 25. Implemented acl_smtp_predata.
5256 26. If the domain in control=submission is set empty, Exim assumes that the
5257 authenticated id is a complete email address when it generates From: or
5258 Sender: header lines.
5260 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5261 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5262 chown and chgrp in /bin and hostname in /usr/bin.
5264 28. Exim was keeping the "process log" file open after each use, just as it
5265 does for the main log. This opens the possibility of it remaining open for
5266 long periods when the USR1 signal hits a daemon. Occasional processlog
5267 errors were reported, that could have been caused by this. Anyway, it seems
5268 much more sensible not to leave this file open at all, so that is what now
5271 29. The long-running daemon process does not normally write to the log once it
5272 has entered its main loop, and it closes the log before doing so. This is
5273 so that log files can straightforwardly be renamed and moved. However,
5274 there are a couple of unusual error situations where the daemon does write
5275 log entries, and I had neglected to close the log afterwards.
5277 30. The text of an SMTP error response that was received during a remote
5278 delivery was being truncated at 512 bytes. This is too short for some of
5279 the long messages that one sometimes sees. I've increased the limit to
5282 31. It is now possible to make retry rules that apply only when a message has a
5283 specific sender, in particular, an empty sender.
5285 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5286 it possible to be selective about when SMTP synchronization is enforced.
5288 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5290 32. Implemented hosts_connection_nolog.
5292 33. Added an ACL for QUIT.
5294 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5297 35. Added mailbox_size and mailbox_filecount to appendfile.
5299 36. Added control = no_multiline_responses to ACLs.
5301 37. There was a bug in the logic of the code that waits for the clock to tick
5302 in the case where the clock went backwards by a substantial amount such
5303 that the microsecond fraction of "now" was more than the microsecond
5304 fraction of "then" (but the whole seconds number was less).
5306 38. Added support for the libradius Radius client library this is found on
5307 FreeBSD (previously only the radiusclient library was supported).
5313 1. When certain lookups returned multiple values in the form name=value, the
5314 quoting of the values was not always being done properly. Specifically:
5315 (a) If the value started with a double quote, but contained no whitespace,
5317 (b) If the value contained whitespace other than a space character (i.e.
5318 tabs or newlines or carriage returns) it was not quoted.
5319 This fix has been applied to the mysql and pgsql lookups by writing a
5320 separate quoting function and calling it from the lookup code. The fix
5321 should probably also be applied to nisplus, ibase and oracle lookups, but
5322 since I cannot test any of those, I have not disturbed their existing code.
5324 2. A hit in the callout cache for a specific address caused a log line with no
5325 reason for rejecting RCPT. Now it says "Previous (cached) callout
5326 verification failure".
5328 3. There was an off-by-one bug in the queryprogram router. An over-long
5329 return line was truncated at 256 instead of 255 characters, thereby
5330 overflowing its buffer with the terminating zero. As well as fixing this, I
5331 have increased the buffer size to 1024 (and made a note to document this).
5333 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5334 when Exim is waiting for an SMTP response from a remote server, Exim
5335 restarts its select() call on the socket, thereby resetting its timeout.
5336 This is not a problem when such interrupts are rare. Somebody set up a cron
5337 job to run exiwhat every 2 minutes, which is less than the normal select()
5338 timeout (5 or 10 minutes). This meant that the select() timeout never
5339 kicked in because it was always reset. I have fixed this by comparing the
5340 time when an interrupt arrives with the time at the start of the first call
5341 to select(). If more time than the timeout has elapsed, the interrupt is
5342 treated as a timeout.
5344 5. Some internal re-factoring in preparation for the addition of Sieve
5345 extensions (by MH). In particular, the "personal" test is moved to a
5346 separate function, and given an option for scanning Cc: and Bcc: (which is
5347 not set for Exim filters).
5349 6. When Exim created an email address using the login of the caller as the
5350 local part (e.g. when creating a From: or Sender: header line), it was not
5351 quoting the local part when it contained special characters such as @.
5353 7. Installed new OpenBSD configuration files.
5355 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5356 try to make them clearer.
5358 9. Callout options, other than the timeout value, were being ignored when
5359 verifying sender addresses in header lines. For example, when using
5361 verify = header_sender/callout=no_cache
5363 the cache was (incorrectly) being used.
5365 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5366 only the Cygwin environment.
5368 11. When return_path_on_delivery was set as a log selector, if different remote
5369 addresses in the same message used different return paths and parallel
5370 remote delivery occurred, the wrong values would sometimes be logged.
5371 (Whenever a remote delivery process finished, the return path value from
5372 the most recently started remote delivery process was logged.)
5374 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5375 lines when AUTH and/or TLS are in use. This is the "received protocol"
5376 field. Exim used to use "asmtp" for authenticated SMTP, without any
5377 indication (in the protocol name) for TLS use. Now it follows the RFC and
5378 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5379 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5380 names appear in log lines as well as in Received: header lines.
5382 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5383 extensions, and comparison tests, and to fix some bugs.
5385 14. Changes to the "personal" filter test:
5387 (1) The test was buggy in that it was just doing the equivalent of
5388 "contains" tests on header lines. For example, if a user's address was
5389 anne@some.where, the "personal" test would incorrectly be true for
5391 To: susanne@some.where
5393 This test is now done by extracting each address from the header in turn,
5394 and checking the entire address. Other tests that are part of "personal"
5395 are now done using regular expressions (for example, to check local parts
5396 of addresses in From: header lines).
5398 (2) The list of non-personal local parts in From: addresses has been
5399 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5400 taken from the Sieve specification recommendations.
5402 (3) If the message contains any header line starting with "List-" it is
5403 treated as non-personal.
5405 (4) The test for "circular" in the Subject: header line has been removed
5406 because it now seems ill-conceived.
5408 15. Minor typos in src/EDITME comments corrected.
5410 16. Installed latest exipick from John Jetmore.
5412 17. If headers_add on a router specified a text string that was too long for
5413 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5414 of string_sprintf() is now avoided.
5416 18. $message_body_size was not set (it was always zero) when running the DATA
5417 ACL and the local_scan() function.
5419 19. For the "mail" command in an Exim filter, no default was being set for
5420 the once_repeat time, causing a random time value to be used if "once" was
5421 specified. (If the value happened to be <= 0, no repeat happened.) The
5422 default is now 0s, meaning "never repeat". The "vacation" command was OK
5423 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5424 (I found it when inspecting the code).
5426 20. There is now an overall timeout for performing a callout verification. It
5427 defaults to 4 times the callout timeout, which applies to individual SMTP
5428 commands during the callout. The overall timeout applies when there is more
5429 than one host that can be tried. The timeout is checked before trying the
5430 next host. This prevents very long delays if there are a large number of
5431 hosts and all are timing out (e.g. when the network connections are timing
5432 out). The value of the overall timeout can be changed by specifying an
5433 additional sub-option for "callout", called "maxwait". For example:
5435 verify = sender/callout=5s,maxwait=20s
5437 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5438 to the end before writing, but this should make it even safer).
5440 22. Exim was forgetting that it had advertised PIPELINING for the second and
5441 subsequent messages on an SMTP connection. It was also not resetting its
5442 memory on STARTTLS and an internal HELO.
5444 23. When Exim logs an SMTP synchronization error within a session, it now
5445 records whether PIPELINING has been advertised or not.
5447 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5448 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5449 rather than long int.
5451 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5453 26. Added the never_mail option to autoreply.
5459 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5460 crash if the getsockname() call failed; this can happen if a connection is
5461 closed very soon after it is established. The problem was simply in the
5462 order in which certain operations were done, causing Exim to try to write
5463 to the SMTP stream before it had set up the file descriptor. The bug has
5464 been fixed by making things happen in the correct order.
5470 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5471 before the connection was closed, thus losing the rejection response.
5473 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5474 some early Solaris releases, but causes trouble in current releases where
5475 socklen_t is defined.
5477 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5480 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5483 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5484 that releasing the top of it at the end releases what was used for sub-
5485 expansions (unless the block got too big). However, discard this block if
5486 the first thing is a variable or header, so that we can use its block when
5487 it is dynamic (useful for very large $message_headers, for example).
5489 6. Lookups now cache *every* query, not just the most recent. A new, separate
5490 store pool is used for this. It can be recovered when all lookup caches are
5491 flushed. Lookups now release memory at the end of their result strings.
5492 This has involved some general refactoring of the lookup sources.
5494 7. Some code has been added to the store_xxx() functions to reduce the amount
5495 of flapping under certain conditions.
5497 8. log_incoming_interface used to affect only the <= reception log lines. Now
5498 it causes the local interface and port to be added to several more SMTP log
5499 lines, for example "SMTP connection from", and rejection lines.
5501 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5503 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5505 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5506 could be overwritten at the end of the current message (or the start of a
5507 new message if it was set in a HELO ACL). The value is now preserved for
5508 the duration of the SMTP connection.
5510 12. If a transport had a headers_rewrite setting, and a matching header line
5511 contained an unqualified address, that address was qualified, even if it
5512 did not match any rewriting rules. The underlying bug was that the values
5513 of the flags that permit the existence of unqualified sender and recipient
5514 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5515 non-local messages, and by -bnq for local messages) were not being
5516 preserved with the message after it was received.
5518 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5519 "next input=" as part of the text comprising the host identity instead of
5520 the correct text. The code was using the same buffer for two different
5521 strings. However, depending on which order the printing function evaluated
5522 its arguments, the bug did not always show up. Under Linux, for example, my
5523 test suite worked just fine.
5525 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5526 doesn't work with some older versions of Perl. It has been changed to "my",
5527 which in any case is probably the better facility to use.
5529 15. A really picky compiler found some instances of statements for creating
5530 error messages that either had too many or two few arguments for the format
5533 16. The size of the buffer for calls to the DNS resolver has been increased
5534 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5535 for addresses that have a lot of PTR records. This alleviates a problem; it
5536 does not fully solve it.
5538 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5539 buffer now truncates the list and logs the incident, which is the same
5540 action as happens when Exim is looking up a host name and its aliases.
5541 Previously in this situation something unpredictable would happen;
5542 sometimes it was "internal error: store_reset failed".
5544 18. If a server dropped the connection unexpectedly when an Exim client was
5545 using GnuTLS and trying to read a response, the client delivery process
5546 crashed while trying to generate an error log message.
5548 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5549 string, for example:
5551 warn message = H1: something\nH2: something
5553 the text was added as a single header line from Exim's point of view
5554 though it ended up OK in the delivered message. However, searching for the
5555 second and subsequent header lines using $h_h2: did not work. This has been
5556 fixed. Similarly, if a system filter added multiple headers in this way,
5557 the routers could not see them.
5559 20. Expanded the error message when iplsearch is called with an invalid key to
5560 suggest using net-iplsearch in a host list.
5562 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5563 ACLs are no longer actually imposed (and a message to that effect is
5566 22. If a "gecos" field in a passwd entry contained escaped characters, in
5567 particular, if it contained a \" sequence, Exim got it wrong when building
5568 a From: or a Sender: header from that name. A second bug also caused
5569 incorrect handling when an unquoted " was present following a character
5570 that needed quoting.
5572 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5573 was not being matched caselessly.
5575 24. Arranged for all hyphens in the exim.8 source to be escaped with
5578 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5579 properly. Recipient callout cache records were still being keyed to include
5580 the sender, even when use_sender was set false. This led to far more
5581 callouts that were necessary. The sender is no longer included in the key
5582 when use_sender is false.
5584 26. Added "control = submission" modifier to ACLs.
5586 27. Added the ${base62d: operator to decode base 62 numbers.
5588 28. dnsdb lookups can now access SRV records.
5590 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5591 the configuration file.
5593 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5594 (-v) mode. This makes the output for a verbose queue run more intelligible.
5596 31. Added a use_postmaster feature to recipient callouts.
5598 32. Added the $body_zerocount variable, containing the number of binary zero
5599 bytes in the message body.
5601 33. The time of last modification of the "new" subdirectory is now used as the
5602 "mailbox time last read" when there is a quota error for a maildir
5605 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5607 35. Added +ignore_unknown as a special item in host lists.
5609 36. Code for decoding IPv6 addresses in host lists is now included, even if
5610 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5611 address was recognized as an IP address, but was then not correctly decoded
5612 into binary, causing unexpected and incorrect effects when compared with
5619 1. Very minor rewording of debugging text in manualroute to say "list of
5620 hosts" instead of "hostlist".
5622 2. If verify=header_syntax was set, and a header line with an unqualified
5623 address (no domain) and a large number of spaces between the end of the
5624 name and the colon was received, the reception process suffered a buffer
5625 overflow, and (when I tested it) crashed. This was caused by some obsolete
5626 code that should have been removed. The fix is to remove it!
5628 3. When running in the test harness, delay a bit after writing a bounce
5629 message to get a bit more predictability in the log output.
5631 4. Added a call to search_tidyup() just before forking a reception process. In
5632 theory, someone could use a lookup in the expansion of smtp_accept_max_
5633 per_host which, without the tidyup, could leave open a database connection.
5635 5. Added the variables $recipient_data and $sender_data which get set from a
5636 lookup success in an ACL "recipients" or "senders" condition, or a router
5637 "senders" option, similar to $domain_data and $local_part_data.
5639 6. Moved the writing of debug_print from before to after the "senders" test
5642 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5643 problems for message scanning, either using a data ACL, or using
5644 local_scan() because the Received: header was not generated till after they
5645 were called (in order to set the time as the time of reception completion).
5646 I have revised the way this works. The header is now generated after the
5647 body is received, but before the ACL or local_scan() are called. After they
5648 are run, the timestamp in the header is updated.
5654 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5655 before starting a queue runner without re-exec. This happened only when
5656 deliver_drop_privilege was set or when the Exim user was set to root. The
5657 effect of the bug was that timeouts during subsequent deliveries caused
5658 crashes instead of being properly handled. The handler is now left at its
5659 default (and expected) setting.
5661 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5662 message, again when deliver_drop_privilege is set or Exim is run as root.
5663 The bug described in (1) was not present in this case, but the tidying up
5664 of the other signals was missing. I have made the two cases consistent.
5666 3. The ignore_target_hosts setting on a manualroute router was being ignored
5667 for hosts that were looked up using the /MX notation.
5669 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5672 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5673 operated on the sender address. After changing the $sender_address to <>
5674 for the sender address verify, Exim was re-instated it as the original
5675 (before rewriting) address, but remembering that it had rewritten it, so it
5676 wasn't rewriting it again. This bug also had the effect of breaking the
5677 sender address verification caching when the sender address was rewritten.
5679 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5680 This has been changed so that if the ip literal address matches
5681 ignore_target_hosts, the router declines.
5683 7. Added expansion conditions match_domain, match_address, and match_local_
5684 part (NOT match_host).
5686 8. The placeholder for the Received: header didn't have a length field set.
5688 9. Added code to Exim itself and to exim_lock to test for a specific race
5689 condition that could lead to file corruption when using MBX delivery. The
5690 issue is with the lockfile that is created in /tmp. If this file is removed
5691 after a process has opened it but before that process has acquired a lock,
5692 there is the potential for a second process to recreate the file and also
5693 acquire a lock. This could lead to two Exim processes writing to the file
5694 at the same time. The added code performs the same test as UW imapd; it
5695 checks after acquiring the lock that its file descriptor still refers to
5696 the same named file.
5698 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5699 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5702 11. Added the smtp_active_hostname option. If used, this will typically be made
5703 to depend on the incoming interface address. Because $interface_address is
5704 not set up until the daemon has forked a reception process, error responses
5705 that can happen earlier (such as "too many connections") no longer contain
5708 12. If an expansion in a condition on a "warn" statement fails because a lookup
5709 defers, the "warn" statement is abandoned, and the next ACL statement is
5710 processed. Previously this caused the whole ACL to be aborted.
5712 13. Added the iplsearch lookup type.
5714 14. Added ident_timeout as a log selector.
5716 15. Added tls_certificate_verified as a log selector.
5718 16. Added a global option tls_require_ciphers (compare the smtp transport
5719 option of the same name). This controls incoming TLS connections.
5721 17. I finally figured out how to make tls_require_ciphers do a similar thing
5722 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5723 before starting the TLS session.
5725 18. Tabs are now shown as \t in -bP output.
5727 19. If the log selector return_path_on_delivery was set, Exim crashed when
5728 bouncing a message because it had too many Received: header lines.
5730 20. If two routers both had headers_remove settings, and the first one included
5731 a superfluous trailing colon, the final name in the first list and the
5732 first name in the second list were incorrectly joined into one item (with a
5733 colon in the middle).
5739 1. Added -C and -D options to the exinext utility, mainly to make it easier
5740 to include in the automated testing, but these could be helpful when
5741 multiple configurations are in use.
5743 2. The exinext utility was not formatting the output nicely when there was
5744 an alternate port involved in the retry record key, nor when there was a
5745 message id as well (for retries that were specific to a specific message
5746 and a specific host). It was also confused by IPv6 addresses, because of
5747 the additional colons they contain. I have fixed the IPv4 problem, and
5748 patched it up to do a reasonable job for IPv6.
5750 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5751 delivery, the log line now contains "pipelined" if PIPELINING was used.
5753 4. An SMTP transport process used to panic and die if the bind() call to set
5754 an explicit outgoing interface failed. This has been changed; it is now
5755 treated in the same way as a connect() failure.
5757 5. A reference to $sender_host_name in the part of a conditional expansion
5758 that was being skipped was still causing a DNS lookup. This no longer
5761 6. The def: expansion condition was not recognizing references to header lines
5762 that used bh_ and bheader_.
5764 7. Added the _cache feature to named lists.
5766 8. The code for checking quota_filecount in the appendfile transport was
5767 allowing one more file than it should have been.
5769 9. For compatibility with Sendmail, the command line option
5777 and sets the incoming protocol and host name (for trusted callers). The
5778 host name and its colon can be omitted when only the protocol is to be set.
5779 Note the Exim already has two private options, -pd and -ps, that refer to
5780 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5781 "s", but I don't think that's a major issue.
5783 10. A number of refactoring changes to the code, none of which should affect
5786 (a) The number of logging options was getting close to filling up the
5787 32-bit word that was used as a bit map. I have split them into two classes:
5788 those that are passed in the argument to log_write(), and those that are
5789 only ever tested independently outside of that function. These are now in
5790 separate 32-bit words, so there is plenty of room for expansion again.
5791 There is no change in the user interface or the logging behaviour.
5793 (b) When building, for example, log lines, the code previously used a
5794 macro that called string_cat() twice, in order to add two strings. This is
5795 not really sufficiently general. Furthermore, there was one instance where
5796 it was actually wrong because one of the argument was used twice, and in
5797 one call a function was used. (As it happened, calling the function twice
5798 did not affect the overall behaviour.) The macro has been replaced by a
5799 function that can join an arbitrary number of extra strings onto a growing
5802 (c) The code for expansion conditions now uses a table and a binary chop
5803 instead of a serial search (which was left over from when there were very
5804 few conditions). Also, it now recognizes conditions like "pam" even when
5805 the relevant support is not compiled in: a suitably worded error message is
5806 given if an attempt is made to use such a condition.
5808 11. Added ${time_interval:xxxxx}.
5810 12. A bug was causing one of the ddress fields not to be passed back correctly
5811 from remote delivery subprocesses. The field in question was not being
5812 subsequently used, so this caused to problems in practice.
5814 13. Added new log selectors queue_time and deliver_time.
5816 14. Might have fixed a bug in maildirsizefile handling that threw up
5817 "unexpected character" debug warnings, and recalculated the data
5818 unnecessarily. In any case, I expanded the warning message to give more
5821 15. Added the message "Restricted characters in address" to the statements in
5822 the default ACL that block characters like @ and % in local parts.
5824 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5825 Three changes have been made:
5827 (a) There was a serious bug; a negative response to MAIL caused the whole
5828 recipient domain to be cached as invalid, thereby blocking all messages
5829 to all local parts at the same domain, from all senders. This bug has
5830 been fixed. The domain is no longer cached after a negative response to
5831 MAIL if the sender used is not empty.
5833 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5836 (c) A new callout option, "use_sender" has been added for people who want
5837 the modified behaviour.
5843 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5846 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5847 indeed breaks things for older releases.
5849 3. Added additional logging to the case where there is a problem reading data
5850 from a filter that is running in a subprocess using a pipe, in order to
5851 try to track down a specific problem.
5853 4. Testing facility fudge: when running in the test harness and attempting
5854 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5855 getting "No route to host". Convert this to a timeout.
5857 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5860 6. Some OS don't have socklen_t but use size_t instead. This affects the
5861 fifth argument of getsockopt() amongst other things. This is now
5862 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5863 can be set for individual OS. I have set it for SunOS5, OSF1, and
5864 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5865 some earlier ones do not.
5867 7. Change 4.30/15 was not doing the test caselessly.
5869 8. The standard form for an IPv6 address literal was being rejected by address
5870 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5871 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5872 this, as well as the form without the "IPv6" on the front (but only when
5873 address literals are enabled, of course).
5875 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5877 10. Exim crashed if a message with an empty sender address specified by -f
5878 encountered a router with an errors_to setting. This could be provoked only
5879 by a command such as
5883 where an empty string was supplied; "<>" did not hit this bug.
5885 11. Installed PCRE release 4.5.
5887 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5888 remained set. It is now erased.
5890 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5891 times from message ids (which are base 36 rather than the normal 62).
5893 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5894 were being counted as actual protocol errors, and logged if the log
5895 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5896 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5897 and DATA following a set of rejected RCPTs do not count as protocol errors.
5898 In other words, Exim assumes they were pipelined, though this may not
5899 actually be the case. Of course, in all cases the client gets an
5900 appropriate error code.
5902 15. If a lookup fails in an ACL condition, a message about the failure may
5903 be available; it is used if testing the ACL cannot continue, because most
5904 such messages specify what the cause of the deferral is. However, some
5905 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5906 that caused an old message to be retained and used if a later statement
5907 caused a defer, replacing the real cause of the deferral.
5909 16. If an IP address had so many PTR records that the DNS lookup buffer
5910 was not large enough to hold them, Exim could crash while trying to process
5911 the truncated data. It now detects and logs this case.
5913 17. Further to 4.21/58, another change has been made: if (and only if) the
5914 first line of a message (the first header line) ends with CRLF, a bare LF
5915 in a subsequent header line has a space inserted after it, so as not to
5916 terminate the header.
5918 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5919 unnecessarily, used atoi() instead of strtol(), and didn't check the
5920 termination when getting file sizes from file names by regex.
5922 19. Completely re-implemented the support for maildirsize files, in the light
5923 of a number of problems with the previous contributed implementation
5924 (4.30/29). In particular:
5926 . If the quota is zero, the maildirsize file is maintained, but no quota is
5929 . If the maildir directory does not exist, it is created before any attempt
5930 to write a maildirsize file.
5932 . The quota value in the file is just a cache; if the quota is changed in
5933 the transport, the new value overrides.
5935 . A regular expression is available for excluding directories from the
5938 20. The autoreply transport checks the characters in options that define the
5939 message's headers; it allows continued headers, but it was checking with
5940 isspace() after an embedded newline instead of explicitly looking for a
5943 21. If all the "regular" hosts to which an address was routed had passed their
5944 expiry times, and had not reached their retry times, the address was
5945 bounced, even if fallback hosts were defined. Now Exim should go on to try
5948 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5949 equivalent code in the SMTP transport. Some hosts send humungous responses
5950 to HELO/EHLO, more than 1024 it seems.
5952 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5953 gives compiler warnings in some environments. I've now done it "properly",
5956 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5957 (because of problems with the built-in one) was declared to return uschar *
5958 instead of char *, causing compiler failure.
5960 25. Fixed a file descriptor leak when processing alias/forward files.
5962 26. Fixed a minor format string issue in dbfn.c.
5964 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5966 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5967 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5970 29. When a sender address is verified, it is cached, to save repeating the test
5971 when there is more than one recipient in a message. However, when the
5972 verification involves a callout, it is possible for different callout
5973 options to be set for different recipients. It is too complicated to keep
5974 track of this in the cache, so now Exim always runs a verification when a
5975 callout is required, relying on the callout cache for the optimization.
5976 The overhead is duplication of the address routing, but this should not be
5979 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5980 to be verified with callout=postmaster, and the main callout worked but the
5981 postmaster check failed, the verification correctly failed. However, if a
5982 subsequent RCPT command asked for sender verification *without* the
5983 postmaster check, incorrect caching caused this verification also to fail,
5986 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5987 it was not caching the DNS options (qualify_single, search_parents) that
5988 were used when the lookup failed. A subsequent lookup with different
5989 options therefore always gave the same answer, though there were cases
5990 where it should not have. (Example: a "domains = !$mx_any" option on a
5991 dnslookup router: the "domains" option is always processed without any
5992 widening, but the router might have qualify_single set.) Now Exim uses the
5993 cached value only when the same options are set.
5995 32. Added John Jetmore's "exipick" utility to the distribution.
5997 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5998 than a timeout (e.g. a certificate is required, and is not provided), an
5999 Exim server now closes the connection immediately. Previously it waited for
6000 the client to close - but if the client is SSL, it seems that they each
6001 wait for each other, leading to a delay before one of them times out.
6003 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6004 maintained 0.8.x compatibility because I don't think many are using it, and
6005 it is clearly obsolete.
6007 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6010 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6011 client certificate was expired. A simple patch fixes this, though I don't
6012 understand the full logic of why the verify callback is called multiple
6015 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6016 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6017 which causes problems with some clients (such as the Certicom SSL Plus
6018 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6019 disables the coutermeasure allowing Eudora to connect."
6021 38. Exim was not checking that a write() to a log file succeeded. This could
6022 lead to Bad Things if a log got too big, in particular if it hit a file
6023 size limit. Exim now panics and dies if it cannot write to a log file, just
6024 as it does if it cannot open a log file.
6026 39. Modified OS/Makefile-Linux so that it now contains
6028 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6030 The two -D definitions ensure that Exim is compiled with large file
6031 support, which makes it possible to handle log files that are bigger than
6034 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6035 instance) a domain was checked against a named list that involved a lookup,
6036 causing $domain_data to be set, then another domain was checked against the
6037 same list, then the first domain was re-checked, the value of $domain_data
6038 after the final check could be wrong. In particular, if the second check
6039 failed, it could be set empty. This bug probably also applied to
6042 41. The strip_trailing_dot option was not being applied to the address given
6043 with the -f command-line option.
6045 42. The code for reading a message's header from the spool was incrementing
6046 $received_count, but never initializing it. This meant that the value was
6047 incorrect (doubled) while delivering a message in the same process in which
6048 it was received. In the most common configuration of Exim, this never
6049 happens - a fresh exec is done - but it can happen when
6050 deliver_drop_privilege is set.
6052 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6053 it now includes up to 150 characters of the unexpected data in the log
6056 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6057 and building data strings. The size of both of these buffers was 10 000
6058 bytes - far larger than anybody would *ever* want, thought I. Needless to
6059 say, somebody hit the limit. I have increased the maximum line length to
6060 20 000 and the maximum data length of concatenated lines to 100 000. I have
6061 also fixed two bugs, because there was no checking on these buffers. Tsk,
6062 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6063 buffer is too small.
6065 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6066 lsearch lookups. Now it does.
6068 46. When parsing a route_list item in a manualroute router, a fixed-length
6069 buffer was used for the list of hosts. I made this 1024 bytes long,
6070 thinking that nobody would ever have a list of hosts that long. Wrong.
6071 Somebody had a whole pile of complicated expansion conditions, and the
6072 string was silently truncated, leading to an expansion error. It turns out
6073 that it is easier to change to an unlimited length (owing to other changes
6074 that have happened since this code was originally written) than to build
6075 structure for giving a limitation error. The length of the item that
6076 expands into the list of hosts is now unlimited.
6078 47. The lsearch lookup could not handle data where the length of text line was
6079 more than 4095 characters. Such lines were truncated, leading to shortened
6080 data being returned. It should now handle lines of any length.
6082 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6083 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6086 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6087 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6088 no longer happens in comment lines. A list of replacements is now placed
6089 at the head of all of the source files, except those whose only change is
6090 to replace PERL_COMMAND in the very first #! line.
6092 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6093 messages on the queue, with a bottom-up merge sort, using code contributed
6094 by Michael Haardt. This should make operations like -bp somewhat faster on
6095 large queues. It won't affect queue runners, except when queue_run_in_order
6098 51. Installed eximstats 1.31 in the distribution.
6100 52. Added support for SRV lookups to the dnslookup router.
6102 53. If an ACL referred to $message_body or $message_body_end, the value was not
6103 reset for any messages that followed in the same SMTP session.
6105 54. The store-handling optimization for building very long strings was not
6106 differentiating between the different store pools. I don't think this
6107 actually made any difference in practice, but I've tidied it.
6109 55. While running the routers to verify a sender address, $sender_address
6110 was still set to the sender address. This is wrong, because when routing to
6111 send a bounce to the sender, it would be empty. Therefore, I have changed
6112 it so that, while verifying a sender address, $sender_address is set to <>.
6113 (There is no change to what happens when verifying a recipient address.)
6115 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6116 target A or AAAA records (if not already returned) without resetting the
6117 qualify_single or search_parents options of the DNS resolver. These are
6118 inappropriate in this case because the targets of MX and SRV records must
6119 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6120 target that, when qualified, matched something in the local domain. These
6121 two options are now turned off when doing these lookups.
6123 57. It seems that at least some releases of Reiserfs (which does not have the
6124 concept of a fixed number of inodes) returns zero and not -1 for the
6125 number of available inodes. This interacted badly with check_spool_inodes,
6126 which assumed that -1 was the "no such thing" setting. What I have done is
6127 to check that the total number of inodes is greater than zero before doing
6128 the test of how many are available.
6130 58. When a "warn" ACL statement has a log_message modifier, the message is
6131 remembered, and not repeated. This is to avoid a lot of repetition when a
6132 message has many recipients that cause the same warning to be written.
6133 However, Exim was preserving the list of already written lines for an
6134 entire SMTP session, which doesn't seem right. The memory is now reset if a
6135 new message is started.
6137 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6138 debugging output unless log_rewrite was also set.
6140 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6141 of (int)(handle) when we know that handle contains (void *)(-1).
6143 61. The Exim daemon panic-logs an error return when it closes the incoming
6144 connection. However "connection reset by peer" seems to be common, and
6145 isn't really an error worthy of noting specially, so that particular error
6148 62. When Exim is trying to find all the local interfaces, it used to panic and
6149 die if the ioctl to get the interface flags failed. However, it seems that
6150 on at least one OS (Solaris 9) it is possible to have an interface that is
6151 included in the list of interfaces, but for which you get a failure error
6152 for this call. This happens when the interface is not "plumbed" into a
6153 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6154 failure of the "get flags" call assumes that the interface is down.
6156 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6157 makes life easier for people who are doing arithmetic on fields extracted
6158 from dates, where you often get leading zeros that should not be
6159 interpreted as octal.
6161 64. Added qualify_domain to the redirect router, to override the global
6164 65. If a pathologically long header line contained very many addresses (the
6165 report of this problem mentioned 10 000) and each of them was rewritten,
6166 Exim could use up a very large amount of memory. (It kept on making new
6167 copies of the header line as it rewrote, and never released the old ones.)
6168 At the expense of a bit more processing, the header rewriting function has
6169 been changed so that it no longer eats memory in this way.
6171 66. The generation of the Received: header has been moved from the time that a
6172 message starts to be received, to the time that it finishes. The timestamp
6173 in the Received: header should now be very close to that of the <= log
6174 line. There are two side-effects of this change:
6176 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6177 logged header lines no longer include the local Received: line, because
6178 it has not yet been created. The same applies to a copy of the message
6179 that is returned to a non-SMTP sender when a message is rejected.
6181 (b) When a filter file is tested using -bf, no additional Received: header
6182 is added to the test message. After some thought, I decided that this
6185 This change does not affect the value of $received_for. It is still set
6186 after address rewriting, but before local_scan() is called.
6188 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6190 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6191 gave an unhelpful panic error message, and a defer error. I have managed to
6192 change this behaviour so that it now rejects any supplied certificate,
6193 which seems right, as the list of acceptable certificates is empty.
6195 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6196 gave an unhelpful defer error. I have not managed to make this reject any
6197 supplied certificates, but the error message it gives is "no certificate
6198 supplied", which is not helpful.
6200 70. exigrep's output now also includes lines that are not associated with any
6201 message, but which match the given pattern. Implemented by a patch from
6202 Martin Sluka, which also tidied up the Perl a bit.
6204 71. Recipient callout verification, like sender verification, was using <> in
6205 the MAIL FROM command. This isn't really the right thing, since the actual
6206 sender may affect whether the remote host accepts the recipient or not. I
6207 have changed it to use the actual sender in the callout; this means that
6208 the cache record is now keyed on a recipient/sender pair, not just the
6209 recipient address. There doesn't seem to be a real danger of callout loops,
6210 since a callout by the remote host to check the sender would use <>.
6211 [SEE ABOVE: changed after hitting problems.]
6213 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6214 temporary errors. However, in the case of such a code being given after
6215 the end of a data transmission (i.e. after ".") Exim was failing to write
6216 a retry record for the message. (Yes, there was some broken host that was
6217 actually sending 8xx at this point.)
6219 73. An unknown lookup type in a host list could cause Exim to panic-die when
6220 the list was checked. (An example that provoked this was putting <; in the
6221 middle of a list instead of at the start.) If this happened during a DATA
6222 ACL check, a -D file could be left lying around. This kind of configuration
6223 error no longer causes Exim to die; instead it causes a defer error. The
6224 incident is still logged to the main and panic logs.
6226 74. Buglet left over from Exim 3 conversion. The message "too many messages
6227 in one connection" was written to the rejectlog but not the mainlog, except
6228 when address rewriting (yes!) was being logged.
6230 75. Added write_rejectlog option.
6232 76. When a system filter was run not as root (that is, when system_filter_user
6233 was set), the values of the $n variables were not being returned to the
6234 main process; thus, they were not subsequently available in the $sn
6237 77. Added +return_path_on_delivery log selector.
6239 78. A connection timeout was being treated differently from recipients deferred
6240 when testing hosts_max_try with a message that was older than the host's
6241 retry timeout. (The host should not be counted, thus allowing all hosts to
6242 be tried at least once before bouncing.) This may have been the cause of an
6243 occasionally reported bug whereby a message would remain on the queue
6244 longer than the retry timeout, but would be bounced if a delivery was
6245 forced. I say "may" because I never totally pinned down the problem;
6246 setting up timeout/retry tests is difficult. See also the next item.
6248 79. The ultimate address timeout was not being applied to errors that involved
6249 a combination of host plus message (for example, a timeout on a MAIL
6250 command). When an address resolved to a number of possible hosts, and they
6251 were not all tried for each delivery (e.g. because of hosts_max_try), a
6252 message could remain on the queue longer than the retry timeout.
6254 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6257 81. Fixed an obscure SMTP outgoing bug which required at least the following
6258 conditions: (a) there was another message waiting for the same server;
6259 (b) the server returned 5xx to all RCPT commands in the first message so
6260 that the message was not completed; (c) the server dropped the connection
6261 or gave a negative response to the RSET that Exim sends to abort the
6262 transaction. The observed case was a dropped connection after DATA that had
6263 been sent in pipelining mode. That is, the server had advertised PIPELINING
6264 but was not implementing it correctly. The effect of the bug was incorrect
6265 behaviour, such as trying another host, and this could lead to a crash.
6271 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6272 and daemon.c were passed as pointers to ints; they should have been
6273 pointers to socklen_t variables (which are typically unsigned ints).
6275 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6278 3. Fixed a really odd bug that affected only the testing scheme; patching a
6279 certain fixed string in the binary changed the value of another string that
6280 happened to be identical to the end of the original first string.
6282 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6283 name", it returns that address as the IP address. On some operating
6284 systems (e.g. Solaris), it also passes back the IP address string as the
6285 "host name". However, on others (e.g. Linux), it passes back an empty
6286 string. Exim wasn't checking for this, and was changing the host name to an
6287 empty string, assuming it had been canonicalized.
6289 5. Although rare, it is permitted to have more than one PTR record for a given
6290 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6291 all the names associated with an address, because they do in Solaris.
6292 However, it seems that they do not in Linux for data that comes from the
6293 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6294 I found this out when I moved to a new Linux workstation and tried to run
6295 the Exim test suite.
6297 To get round this problem I have changed the code so that it now does its
6298 own call to the DNS to look up PTR records when searching for a host name.
6299 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6300 addresses that are only in /etc/hosts are still found.
6302 This behaviour is, however, controlled by an option called host_lookup_
6303 order, which defaults to "bydns:byaddr". If people want to use the other
6304 order, or indeed, just use one or the other means of lookup, they can
6305 specify it in this variable.
6307 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6308 some operating systems, this comes back from gethostbyaddr() as an empty
6309 string, and this is what Exim used to test for. However, it seems that in
6310 other systems, "." is yielded. Exim now tests for this case too.
6312 7. The values of check_spool_space and check_log_space are now held internally
6313 as a number of kilobytes instead of an absolute number of bytes. If a
6314 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6315 kilobyte. This means that much larger values can be stored.
6317 8. Exim monitor: an attempt to get the action menu when not actually pointing
6318 at a message produces an empty menu entitled "No message selected". This
6319 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6320 no entries in it ("Shell widget menu has zero width and/or height"). So I
6321 have added a single, blank menu entry in this case.
6323 9. Added ${quote_local_part.
6325 10. MIME decoding is now applied to the contents of Subject: header lines when
6328 11. Now that a reference to $sender_host_address automatically causes a reverse
6329 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6330 host lookup before query-style lookups in lists that might use this
6331 variable. This has therefore been abolished, and the "net-" prefix is no
6332 longer necessary for query-style lookups.
6334 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6335 have been a typo for LFLAGS, so it has been changed.
6337 13. The install script calls Exim with "-C /dev/null" in order to find the
6338 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6339 to be output. However, since Exim outputs its version number before the
6340 error, it didn't break the script. It just looked ugly. I fixed this by
6341 always allowing "-C /dev/null" if the caller is root.
6343 14. Ignore overlarge ACL variable number when reading spool file - insurance
6344 against a later release with more variables having written the file.
6346 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6347 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6348 this, as well as the form without the "IPv6" on the front.
6350 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6351 OS/Makefile-Darwin file.
6353 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6354 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6356 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6357 variable. However, some OS use ints for inodes. I've added cast to long int
6358 to get rid of the compiler warning.
6360 19. I had forgotten to lock out "/../" in configuration file names when
6361 ALT_CONFIG_PREFIX was set.
6363 20. Routers used for verification do not need to specify transports. However,
6364 if such a router generated a host list, and callout was configured, Exim
6365 crashed, because it could not find a port number from the (non-existent)
6366 transport. It now assumes port 25 in this circumstance.
6368 21. Added the -t option to exigrep.
6370 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6371 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6372 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6373 them is set, LOOKUP_LSEARCH is forced.
6375 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6377 24. Added sender and host information to the "rejected by local_scan()" log
6378 line; previously there was no indication of these.
6380 25. Added .include_if_exists.
6382 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6383 Linux. It turns out that not all file systems support this. Apparently some
6384 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6385 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6386 not supported on the file descriptor, is now ignored when Exim is trying to
6387 sync a directory. This applies only to Linux.
6389 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6391 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6392 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6395 29. Added support for maildirsize files from supplied patch (modified a bit).
6397 30. The use of :fail: followed by an empty string could lead Exim to respond to
6398 sender verification failures with (e.g.):
6400 550 Verification failed for <xxx>
6401 550 Sender verify failed
6403 where the first response line was missing the '-' that indicates it is not
6404 the final line of the response.
6406 31. The loop for finding the name of the user that called Exim had a hardwired
6407 limit of 10; it now uses the value of finduser_retries, which is used for
6408 all other user lookups.
6410 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6413 33. Exim was neglecting to zero errno before one call of strtol() when
6414 expanding a string and expecting an integer value. On some systems this
6415 resulted in spurious "integer overflow" errors. Also, it was casting the
6416 result into an int without checking.
6418 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6419 did not work. The code looks as if it has *never* worked, though it appears
6420 to have been documented since at least release 1.62. I have made it work.
6422 35. The "timeout_DNS" error in retry rules, also documented since at least
6423 1.62, also never worked. As it isn't clear exactly what this means, and
6424 clearly it isn't a major issue, I have abolished the feature by treating it
6425 as "timeout", and writing a warning to the main and panic logs.
6427 36. The display of retry rules for -brt wasn't always showing the error code
6430 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6431 timeout_connect_A, timeout_connect_MX.
6433 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6434 to the empty sender.
6436 39. The daemon was not analysing the content of -oX till after it had closed
6437 stderr and disconnected from the controlling terminal. This meant that any
6438 syntax errors were only noted on the panic log, and the return code from
6439 the command was 0. By re-arranging the code a little, I've made the
6440 decoding happen first, so such errors now appear on stderr, and the return
6441 code is 1. However, the actual setting up of the sockets still happens in
6442 the disconnected process, so errors there are still only recorded on the
6445 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6446 connections (as happens on some IP stacks) was logged at start up time as
6447 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6448 it from "IPv6 and IPv4", which means that two separate sockets are being
6451 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6452 says whether AF_INET or AF_INET6 was passed as an argument.
6454 42. Exiwhat output was messed up when time zones were included in log
6457 43. Exiwhat now gives more information about the daemon's listening ports,
6458 and whether -tls-on-connect was used.
6460 44. The "port" option of the smtp transport is now expanded.
6462 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6463 silently ignored. Now an error message is written to the main and panic
6466 46. There's a new ACL modifier called "logwrite" which writes to a log file
6467 as soon as it is encountered.
6469 47. Added $local_user_uid and $local_user_gid at routing time.
6471 48. Exim crashed when trying to verify a sender address that was being
6474 49. Exim was recognizing only a space character after ".include". It now also
6475 recognizes a tab character.
6477 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6478 extracting the relevant information from the specification. The man page no
6479 longer contains scrambled data for the -d option, and I've added a section
6480 at the front about calling Exim under different names.
6482 51. Added "extra_headers" argument to the "mail" command in filter files.
6484 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6487 53. Installed eximstats 1.29.
6489 54. Added transport_filter_timeout as a generic transport option.
6491 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6492 Cc: header lines. This was required by RFC 822, but it not required by RFC
6495 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6496 incoming messages that did not have them. Now it does so only if the
6497 message originates locally, that is, if there is no associated remote host
6498 address. When Resent- header lines are present, this applies to the Resent-
6499 lines rather than the non-Resent- lines.
6501 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6502 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6504 58. Messages for configuration errors now include the name of the main
6505 configuration file - useful now that there may be more than one file in a
6506 list (.included file names were always shown).
6508 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6509 for those rare installations that do not start the daemon as root or run it
6510 setuid root. I've cut out the call to initgroups() if the daemon is not
6513 60. The Exim user and group can now be bound into the binary as text strings
6514 that are looked up at the start of Exim's processing.
6516 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6518 62. Added $mailstore_basename variable.
6520 63. Installed patch to sieve.c from Michael Haardt.
6522 64. When Exim failed to open the panic log after failing to open the main log,
6523 the original message it was trying to log was written to stderr and debug
6524 output, but if they were not available (the usual case in production), it
6525 was lost. Now it is written to syslog before the two lines that record the
6526 failures to open the logs.
6528 65. Users' Exim filters run in subprocesses under the user's uid. It is
6529 possible for a "deliver" command or an alias in a "personal" command to
6530 provoke an address rewrite. If logging of address rewriting is configured,
6531 this fails because the process is not running as root or exim. There may be
6532 a better way of dealing with this, but for the moment (because 4.30 needs
6533 to be released), I have disabled address rewrite logging when running a
6534 filter in a non-root, non-exim process.
6540 1. The buildconfig auxiliary program wasn't quoting the value set for
6541 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6542 not defined. This bug was masked in 4.22 by the effect that was fixed in
6545 2. Some messages that were rejected after a message id was allocated were
6546 shown as "incomplete" by exigrep. It no longer does this for messages that
6547 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6549 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6550 have allow_domain_literals set, the ID did not get logged in the <= line.
6551 Domain literals are now always recognized in Message-ID: header lines.
6553 4. The first argument for a ${extract expansion item is the key name or field
6554 number. Leading and trailing spaces in this item were not being ignored,
6555 causing some misleading effects.
6557 5. When deliver_drop_privilege was set, single queue runner processes started
6558 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6559 same command in the process it spins off) were not dropping privilege.
6561 6. When the daemon running as "exim" started a queue runner, it always
6562 re-executed Exim in the spun-off process. This is a waste of effort when
6563 deliver_drop_privilege is set. The new process now just calls the
6564 queue-runner function directly.
6570 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6573 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6574 ignored. Though the use of -oP was forcing the writing of a pid file, it
6575 was always written to the default place.
6577 3. If the message "no IP address found for host xxxx" is generated during
6578 incoming verification, it is now followed by identification of the incoming
6579 connection (so you can more easily find what provoked it).
6581 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6583 5. Added some features to "harden" Exim a bit more against certain attacks:
6585 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6586 be put in Local/Makefile. This is like the never_users runtime option,
6587 but it cannot be overridden. The default setting is "root".
6589 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6590 prefix string with which any file named in a -C command line option
6593 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6594 is retained for -C and -D only if the caller of Exim is root. Without
6595 it, the exim user may also use -C and -D and retain privilege.
6597 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6598 command line option is disabled.
6600 6. Macro names set by the -D option must start with an upper case letter, just
6601 like macro names defined in the configuration file.
6603 7. Added "dereference=" facility to LDAP.
6605 8. Two instances of the typo "uknown" in the source files are fixed.
6607 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6608 the Configure-Makefile script screwed up while processing it.
6610 10. Incorporated PCRE 4.4.
6612 11. The SMTP synchronization check was not operating right at the start of an
6613 SMTP session. For example, it could not catch a HELO sent before the client
6614 waited for the greeting. There is now a check for outstanding input at the
6615 point when the greeting is written. Because of the duplex, asynchronous
6616 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6617 way, but not yet received, when the check is performed.
6619 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6620 on TCP/IP sockets, because this apparently causes some broken clients to
6623 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6624 unchanged) from the Cygwin maintainer.
6626 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6627 was supported instead of testing for mbx. Effectively a typo.
6629 15. The spa authenticator server code was not checking that the input it
6630 received was valid base64.
6632 16. The debug output line for the "set" modifier in ACLs was not showing the
6633 name of the variable that was being set.
6635 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6637 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6638 lookup on its own. Something else was needed to trigger the lookup. For
6639 example, a match in host_lookup or the need for a host name in a host list.
6640 Now, if $sender_host_name is referenced and the host name has not yet been
6641 looked up, a lookup is performed. If the lookup fails, the variable remains
6642 empty, and $host_lookup_failed is set to "1".
6644 19. Added "eqi" as a case-independent comparison operator.
6646 20. The saslauthd authentication condition could segfault if neither service
6647 nor realm was specified.
6649 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6650 error message that was logged was misleading, and incoming SMTP
6651 connections were dropped. The message is now more accurate, and temporary
6652 errors are given to SMTP connections.
6654 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6655 (except RSET) with a 421 error, until QUIT is received. However, it was
6656 failing to send a response to QUIT.
6658 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6659 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6660 ACL is now run after the helo_try_verify_hosts code.
6662 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6663 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6664 case-independent, so other case variants are also recognized). Apparently
6665 some systems use these upper case variants.
6667 25. If more than two messages were waiting for the same host, and a transport
6668 filter was specified for the transport, Exim sent two messages over the
6669 same TCP/IP connection, and then failed with "socket operation on non-
6670 socket" when it tried to send the third.
6672 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6674 27. The extern definition of crypt16() in expand.c was not being excluded when
6675 the OS had its own crypt16() function.
6677 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6678 as a preferred synonym for return_size_limit, both as an option and as an
6681 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6683 30. Changed the default configuration ACL to relax the local part checking rule
6684 for addresses that are not in any local domains. For these addresses,
6685 slashes and pipe symbols are allowed within local parts, but the sequence
6686 /../ is explicitly forbidden.
6688 31. SPA server authentication was not clearing the challenge buffer before
6691 32. log_message in a "warn" ACL statement was writing to the reject log as
6692 well as to the main log, which contradicts the documentation and doesn't
6693 seem right (because no rejection is happening). So I have stopped it.
6695 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6696 However, I am unable to do any testing of this.
6698 34. Fixed an infelicity in the appendfile transport. When checking directories
6699 for a mailbox, to see if any needed to be created, it was accidentally
6700 using path names with one or more superfluous leading slashes; tracing
6701 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6703 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6704 contents are added to the log line that is written for every discarded
6705 recipient. (Previously a log_message setting was ignored.)
6707 36. The ${quote: operator now quotes the string if it is empty.
6709 37. The install script runs exim in order to find its version number. If for
6710 some reason other than non-existence or emptiness, which it checks, it
6711 could not run './exim', it was installing it with an empty version number,
6712 i.e. as "exim-". This error state is now caught, and the installation is
6715 38. An argument was missing from the function that creates an error message
6716 when Exim fails to connect to the socket for saslauthd authentication.
6717 This could cause Exim to crash, or give a corrupted message.
6719 39. Added isip, isip4, and isip6 to ${if conditions.
6721 40. The ACL variables $acl_xx are now saved with the message, and can be
6722 accessed later in routers, transports, and filters.
6724 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6725 strings in the file are not string-expanded.
6727 42. If a MAIL command specified a SIZE value that was too large to fit into an
6728 int variable, the check against message_size_limit failed. Such values are
6729 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6730 day this will have to be increased, but I don't think I want to be around
6731 when emails are that large.
6738 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6739 iconv() is not standard in FreeBSD.
6741 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6742 IPv6 enabled. The observed symptom was a segmentation fault on return from
6743 the function os_common_find_running_interfaces() in src/os.c.
6745 3. In the check_special_case() function in daemon.c I had used "errno" as an
6746 argument name, which causes warnings on some systems. This was basically a
6747 typo, since it was named "eno" in the comments!
6749 4. The code that waits for the clock to tick (at a resolution of some fraction
6750 of a second) so as to ensure message-id uniqueness was always waiting for
6751 at least one whole tick, when it could have waited for less. [This is
6752 almost certainly not relevant at current processor speeds, where it is
6753 unlikely to ever wait at all. But we try to future-proof.]
6755 5. The function that sleeps for a time interval that includes fractions of a
6756 second contained a race. It did not block SIGALRM between setting the
6757 timer, and suspending (a couple of lines later). If the interval was short
6758 and the sigsuspend() was delayed until after it had expired, the suspension
6759 never ended. On busy systems this could lead to processes getting stuck for
6762 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6763 process, before it forks any delivery processes. The open lookup caching
6764 mechanism meant that the open file or database connection was passed into
6765 the delivery process. The problem was that delivery processes always tidy
6766 up cached lookup data. This could cause a problem for the next delivery
6767 process started by the queue runner, because the external queue runner
6768 process does not know about the closure. So the next delivery process
6769 still has data in the lookup cache. In the case of a file lookup, there was
6770 no problem because closing a file descriptor in a subprocess doesn't affect
6771 the parent. However, if the lookup was caching a connection to a database,
6772 the connection was closed, and the second delivery process was likely to
6773 see errors such as "PGSQL: query failed: server closed the connection
6774 unexpectedly". The problem has been fixed by closing all cached lookups
6775 in a queue runner before running a delivery process.
6777 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6778 seem to have the "const" qualifier which it has on other OS. I've
6781 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6782 *of the same type* (client or server) with the same public name that an
6783 error should be diagnosed.
6785 9. When Exim looked up a host name for an IP address, but failed to find the
6786 original IP address when looking up the host name (a safety check), it
6787 output the message "<ip address> does not match any IP for NULL", which was
6788 confusing, to say the least. The bug was that the host name should have
6789 appeared instead of "NULL".
6791 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6792 user that is built into the binary, and the -C or -D options is used, root
6793 privilege is dropped before the configuration file is read. In addition,
6794 logging is switched to stderr instead of the normal log files. If the
6795 configuration then re-defines the Exim user, the unprivileged environment
6796 is probably not what is expected, so Exim logs a panic warning message (but
6799 However, if deliver_drop_privilege is set, the unprivileged state may well
6800 be exactly what is intended, so the warning has been cut out in that case,
6801 and Exim is allowed to try to write to its normal log files.
6807 1. smtp_return_error_details was not giving details for temporary sender
6808 or receiver verification errors.
6810 2. Diagnose a configuration error if two authenticators have the same public
6813 3. Exim used not to create the message log file for a message until the first
6814 delivery attempt. This could be confusing when incoming messages were held
6815 for policy or load reasons. The message log file is now created at the time
6816 the message is received, and an initial "Received" line is written to it.
6818 4. The automatically generated man page for command line options had a minor
6819 bug that caused no ill effects; however, a more serious problem was that
6820 the procedure for building the man page automatically didn't always
6821 operate. Consequently, release 4.20 contains an out-of-date version. This
6822 shouldn't happen again.
6824 5. When building Exim with embedded Perl support, the script that builds the
6825 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6826 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6828 6. The freeze_tell option was not being used for messages that were frozen on
6829 arrival, either by an ACL or by local_scan().
6831 7. Added the smtp_incomplete_transaction log selector.
6833 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6834 was accepting AUTH without a new EHLO.
6836 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6837 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6840 10. Logging of TCP/IP connections (when configured) now happens in the main
6841 daemon process instead of the child process, so that the TCP/IP connection
6842 count is more accurate (but it can never be perfect).
6844 11. The use of "drop" in a nested ACL was not being handled correctly in the
6845 outer ACL. Now, if condition failure induced by the nested "drop" causes
6846 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6847 or "require"), the connection is dropped.
6849 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6850 that yield "discard" can now be used with an "accept" or a "discard" verb,
6851 but an error is generated for any others (because I can't see a useful way
6852 to define what should happen).
6854 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6855 are now processed in the same way as lines in the Exim configuration file.
6856 In particular, continuation lines are supported.
6858 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6860 15. Added -ti meaning -t -i.
6862 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6863 domains, and warn by logging if others are found.
6865 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6866 returned from GIFCONF when Exim is trying to find the list of interfaces on
6867 a host. The code in os.c has been modified to copy each ifreq to an aligned
6868 structure in all cases.
6870 Also, in some cases, the returned ifreq's were being copied to a 'struct
6871 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6872 means the last couple of bytes of an IPv6 address could be chopped if the
6873 ifreq contained only a normal sockaddr (14 bytes storage).
6875 18. Named domain lists were not supported in the hosts_treat_as_local option.
6876 An entry such as +xxxx was not recognized, and was treated as a literal
6879 19. Ensure that header lines added by a DATA ACL are included in the reject log
6880 if the ACL subsequently rejects the message.
6882 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6883 MD5 (which is deprecated).
6885 21. When testing a filter file using -bf, Exim was writing a message when it
6886 took the sender from a "From " line in the message, but it was not doing so
6887 when it took $return_path from a Return-Path: header line. It now does.
6889 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6890 with a valid header line field name (a series of printing characters
6891 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6893 23. Changed "disc" in the source to "disk" to conform to the documentation and
6894 the book and for uniformity.
6896 24. Ignore Sendmail's -Ooption=value command line item.
6898 25. When execve() failed while trying to run a command in a pipe transport,
6899 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6900 could be confused with a return value of 69 from the command itself. This
6901 has been changed to 127, the value the shell returns if it is asked to run
6902 a non-existent command. The wording for the related log line suggests a
6903 non-existent command as the problem.
6905 26. If received_header_text expands to an empty string, do not add a Received:
6906 header line to the message. (Well, it adds a token one on the spool, but
6907 marks it "old" so that it doesn't get used or transmitted.)
6909 27. Installed eximstats 1.28 (addition of -nt option).
6911 28. There was no check for failure on the call to getsockname() in the daemon
6912 code. This can fail if there is a shortage of resources on the system, with
6913 ENOMEM, for example. A temporary error is now given on failure.
6915 29. Contrary to the C standard, it seems that in some environments, the
6916 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6917 program. Exim now does this explicitly; it affects the formatting of
6918 timestamps using strftime().
6920 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6921 complaints. I've now initialized all the variables, to avoid this.
6923 32. Header lines added by a system filter were not being "seen" during
6924 transport-time rewrites.
6926 33. The info_callback() function passed to OpenSSL is set up with type void
6927 (*)(SSL *, int, int), as described somewhere. However, when calling the
6928 function (actually a macro) that sets it up, the type void(*)() is
6929 expected. I've put in a cast to prevent warnings from picky compilers.
6931 34. If a DNS black list lookup found a CNAME record, but there were no A
6932 records associated with the domain it pointed at, Exim crashed.
6934 35. If a DNS black list lookup returned more than one A record, Exim ignored
6935 all but the first. It now scans all returned addresses if a particular IP
6936 value is being sought. In this situation, the contents of the
6937 $dnslist_value variable are a list of all the addresses, separated by a
6940 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6941 to accept a host name and all its aliases if the forward lookup for any of
6942 them yielded the IP address of the incoming connection. Now it accepts only
6943 those names whose forward lookup yields the correct IP address. Any other
6944 names are discarded. This closes a loophole whereby a rogue DNS
6945 administrator could create reverse DNS records to break through a
6946 wildcarded host restriction in an ACL.
6948 37. If a user filter or a system filter that ran in a subprocess used any of
6949 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6950 the wrong values were passed to the pipe command ($thisaddress had the
6951 value of $0, $0 had the value of $1, etc). This bug was introduced by
6952 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6954 38. Improved the line breaking for long SMTP error messages from ACLs.
6955 Previously, if there was no break point between 40 and 75 characters, Exim
6956 left the rest of the message alone. Two changes have been made: (a) I've
6957 reduced the minimum length to 35 characters; (b) if it can't find a break
6958 point between 35 and 75 characters, it looks ahead and uses the first one
6959 that it finds. This may give the occasional overlong line, but at least the
6960 remaining text gets split now.
6962 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6963 file descriptors might be low, and that setting 1000 would always raise it.
6964 It turns out that in some environments, the limit is already over 1000 and
6965 that lowering it causes trouble. So now Exim takes care not to decrease it.
6967 40. When delivering a message, the value of $return_path is set to $sender_
6968 address at the start of routing (routers may change the value). By an
6969 oversight, this default was not being set up when an address was tested by
6970 -bt or -bv, which affected the outcome if any router or filter referred to
6973 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6974 log only when "message" or "log_message" are set. However, if one of the
6975 conditions was an address verification, or a call to a nested ACL, the
6976 messages generated by the underlying test were being passed through. This
6977 no longer happens. The underlying message is available in $acl_verify_
6978 message for both "message" and "log_message" expansions, so it can be
6979 passed through if needed.
6981 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6982 new expansion $bh_ to give the encoded byte string without charset
6983 translation. Translation happens only if iconv() is available; HAVE_ICONV
6984 indicates this at build time. HEADERS_CHARSET gives the charset to
6985 translate to; headers_charset can change it in the configuration, and
6986 "headers charset" can change it in an individual filter file.
6988 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6989 that creates RFC 2047 encoded "words" labels them as that charset instead
6990 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6991 expansion operator; (ii) when Exim creates a From: line for a local
6992 message; (iii) when a header line is rewritten to include a "phrase" part.
6994 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6995 buggy, causing it to skip the first lines of messages whose message ID
6996 ended in 'D'. This would not have bitten before Exim release 4.14, because
6997 message IDs were unlikely to end in 'D' before then. The effect was to have
6998 incorrect size information for certain domains.
7000 45. #include "config.h" was missing at the start of the crypt16.c module. This
7001 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7004 46. If there was a timeout during a "random" callout check, Exim treated it as
7005 a failure of the random address, and carried on sending RSET and the real
7006 address. If the delay was just some slowness somewhere, the response to the
7007 original RCPT would be taken as a response to RSET and so on, causing
7008 mayhem of various kinds.
7010 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7011 when I implemented it. It didn't allow for the fact that some option values
7012 may legitimately be negative (e.g. size_addition), and it didn't even do
7013 the right test for positive values.
7015 48. Domain names in DNS records are case-independent. Exim always looks them up
7016 in lower case. Some resolvers return domain names in exactly the case they
7017 appear in the zone file, that is, they may contain uppercase letters. Not
7018 all resolvers do this - some return always lower case. Exim was treating a
7019 change of case by a resolver as a change of domain, similar to a widening
7020 of a domain abbreviation. This triggered its re-routing code and so it was
7021 trying to route what was effectively the same domain again. This normally
7022 caused routing to fail (because the router wouldn't handle the domain
7023 twice). Now Exim checks for this case specially, and just changes the
7024 casing of the domain that it ultimately uses when it transmits the message
7027 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7030 50. If a filter generated a file delivery with a non-absolute name (possible if
7031 no home directory exists for the router), the forbid_file option was not
7034 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7035 the existing equality matching.
7037 52. Exim was using ints instead of ino_t variables in some places where it was
7038 dealing with inode numbers.
7040 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7041 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7042 called TMPDIR, and if it finds it is different, it changes its value.
7044 54. The smtp_printf() function is now made available to local_scan() so
7045 additional output lines can be written before returning. There is also an
7046 smtp_fflush() function to enable the detection of a dropped connection.
7047 The variables smtp_input and smtp_batched_input are exported to
7050 55. Changed the default runtime configuration: the message "Unknown user"
7051 has been removed from the ACL, and instead placed on the localuser router,
7052 using the cannot_route_message feature. This means that any verification
7053 failures that generate their own messages won't get overridden. Similarly,
7054 the "Unrouteable address" message that was in the ACL for unverifiable
7055 relay addresses has also been removed.
7057 56. Added hosts_avoid_esmtp to the smtp transport.
7059 57. The exicyclog script was not checking for the esoteric option
7060 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7061 will work only if exicyclog is run under the appropriate euid.
7063 58. Following a discussion on the list, the rules by which Exim recognises line
7064 endings on incoming messages have been changed. The -dropcr and drop_cr
7065 options are now no-ops, retained only for backwards compatibility. The
7066 following line terminators are recognized: LF CRLF CR. However, special
7067 processing applies to CR:
7069 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7070 nor a local message in the state where . is a terminator.
7072 (ii) If a bare CR is encountered in a header line, an extra space is added
7073 after the line terminator so as not to end the header. The reasoning
7074 behind this is that bare CRs in header lines are most likely either
7075 to be mistakes, or people trying to play silly games.
7077 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7078 was being incorrectly given as 18 bytes larger than it should have been.
7079 This is a VOB (very old bug).
7081 60. This may never have affected anything current, but just in case it has:
7082 When the local host is found other than at the start of a list of hosts,
7083 the local host, those with the same MX, and any that follow, are discarded.
7084 When the list in question was part of a longer list of hosts, the following
7085 hosts (not currently being processed) were also being discarded. This no
7086 longer happens. I'm not sure if this situation could ever has previously
7089 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7092 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7093 header. This is something that is recommended in a new Internet Draft, and
7094 is something that is documented as being done by Sendmail. There are two
7095 possible values. For messages generated by the autoreply transport, Exim
7098 Auto-Submitted: auto-replied
7100 whereas for all other generated messages (e.g. bounces) it adds
7102 Auto-Submitted: auto-generated
7104 63. The "personal" condition in filters now includes a test for the
7105 Auto-Submitted: header. If it contains the string "auto-" the message it
7106 not considered personal.
7108 64. Added rcpt_include_affixes as a generic transport option.
7110 65. Added queue_only_override (default true).
7112 66. Added the syslog_duplication option.
7114 67. If what should have been the first header line of a message consisted of
7115 a space followed by a colon, Exim was mis-interpreting it as a header line.
7116 It isn't of course - it is syntactically invalid and should therefore be
7117 treated as the start of the message body. The misbehaviour could have
7118 caused a number of strange effects, including loss of data in subsequent
7119 header lines, and spool format errors.
7121 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7122 client host had authenticated. This control can now be exercised by an ACL
7123 for more flexibility.
7125 69. By default, callouts do not happen when testing with -bh. There is now a
7126 variant, -bhc, which does actually run the callout code, including
7127 consulting and updating the callout cache.
7129 70. Added support for saslauthd authentication, courtesy of Alexander
7132 71. If statvfs() failed on the spool or log directories while checking their
7133 size for availability, Exim confusingly gave the error "space shortage".
7134 Furthermore, in debugging mode it crashed with a floating point exception.
7135 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7136 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7137 serious problem, Exim now writes to the main and panic logs when this
7138 happens, with details of the failure. It then refuses to accept the
7139 incoming message, giving the message "spool directory problem" or "log
7140 directory problem" with a 421 code for SMTP messages.
7142 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7143 0, 1, and 2 exist, because some OS complain for execs without them (see
7144 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7145 descriptors. However, the code omitted to check that the open succeeded,
7146 causing mysterious errors if for some reason the permissions on /dev/null
7147 got screwed. Now Exim writes a message to the main and panic logs, and
7148 bombs out if it can't open /dev/null.
7150 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7151 interact so that it is all more flexible. It is supposed to remain
7152 backwards compatible. Also added extra_local_interfaces.
7154 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7155 to bomb out with an assertion failure - to the client this appears as a
7156 connection drop. This problem occurs in the part of the code that was taken
7157 from the Samba project. Fortunately, the assertion is in a very simple
7158 function, so I have fixed this by reproducing the function inline in the
7159 one place where it is called, and arranging for authentication to fail
7160 instead of killing the process with assert().
7162 75. The SPA client code was not working when the server requested OEM rather
7163 than Unicode encoding.
7165 76. Added code to make require_files with a specific uid setting more usable in
7166 the case where statting the file as root fails - usually a non-root-mounted
7167 NFS file system. When this happens and the failure is EACCES, Exim now
7168 forks a subprocess and does the per-uid checking as the relevant uid.
7170 77. Added process_log_path.
7172 78. If log_file_path was not explicitly set, a setting of check_log_space or
7173 check_log_inodes was ignored.
7175 79. If a space check for the spool or log partitions fails, the incident is now
7176 logged. Of course, in the latter case the data may get lost...
7178 80. Added the %p formatting code to string_format() so that it can be used to
7179 print addresses in debug_print(). Adjusted all the address printing in the
7180 debugging in store.c to use %p rather than %d.
7182 81. There was a concern that a line of code in smtp_in.c could overflow a
7183 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7184 initially expressed, the concern was not well-founded, because trailing
7185 spaces are removed early. However, if the trailing spaces were followed by
7186 a NULL, they did not get removed, so the overflow was possible. Two fixes
7189 (a) I re-wrote the offending code in a cleaner fashion.
7190 (b) If an incoming SMTP command contains a NULL character, it is rejected
7193 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7194 runs initgroups(), so that if the Exim user is in any additional groups,
7195 they will be used during message reception.
7201 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob